Midline and off‐midline wound closure methods after surgical treatment for pilonidal sinus

Abstract

Background

Pilonidal sinus disease is a common and debilitating condition. Surgical treatment remains the mainstay for managing chronic disease, with options including midline and off‐midline wound closure methods. However, the optimal approach remains uncertain. Recent developments in tension‐free midline techniques require further exploration.

Objectives

To assess the effects of midline and off‐midline wound closure methods for pilonidal sinus, and to determine the optimal off‐midline flap procedures.

Search methods

In June 2022, we searched the Cochrane Wounds Specialised Register, CENTRAL, MEDLINE, Embase, CINAHL Plus EBSCO, and clinical trials registries. We also scanned the reference lists of included studies, as well as reviews, meta‐analyses, and health technology reports. We applied no language, publication date, or study setting restrictions.

Selection criteria

We included parallel RCTs involving participants undergoing midline closure without flap techniques and off‐midline closure for pilonidal sinus treatment. We excluded quasi‐experimental studies and studies that enroled participants presenting with an abscess.

Data collection and analysis

We followed standard Cochrane methodology. The critical outcomes included wound healing (time to wound healing, proportion of wounds healed), recurrence rate, wound infection, wound dehiscence, time to return to work, and quality of life. We assessed biases in these outcomes utilising the Cochrane risk of bias 2 tool and appraised evidence certainty via the GRADE approach.

Main results

We included 33 studies with 3667 analysed participants. The median or average age of the participants across the included studies ranged from 21.0 to 34.2 years, with a predominant male representation. Geographically, the trials were primarily conducted in the Middle East. We identified nine intervention comparisons. In this abstract, we focus on and present the summarised findings for the three primary comparisons.

Off‐midline closure versus conventional midline closure

Off‐midline closure probably reduces the time to wound healing (mean difference (MD) ‐5.23 days, 95% confidence interval (CI) ‐7.55 to ‐2.92 days; 3 studies, 300 participants; moderate‐certainty evidence). However, there may be little to no difference between the two methods in the proportion of wounds healed (100% versus 88.5%, risk ratio (RR) 1.13, 95% CI 0.92 to 1.39; 2 studies, 207 participants; very low‐certainty evidence). Off‐midline closure probably results in lower rates of recurrence (1.5% versus 6.8%, RR 0.22, 95% CI 0.11 to 0.45; 13 studies, 1492 participants; moderate‐certainty evidence) and wound infection (3.8% versus 11.7%, RR 0.32, 95% CI 0.22 to 0.49; 13 studies, 1568 participants; moderate‐certainty evidence), and may lower rates of wound dehiscence (3.9% versus 8.9%, RR 0.44, 95% CI 0.27 to 0.71; 11 studies, 1389 participants; low‐certainty evidence). Furthermore, off‐midline closure may result in a reduced time to return to work (MD ‐3.72 days, 95% CI ‐6.11 to ‐1.33 days; 6 studies, 820 participants; low‐certainty evidence). There were no data available for quality of life.

Off‐midline closure versus tension‐free midline closure

Off‐midline closure may reduce the time to wound healing (median 14 days in off‐midline closure versus 51 days in tension‐free midline closure; 1 study, 116 participants; low‐certainty evidence) and increase wound healing rates at three months (94.7% versus 76.4%, RR 1.24, 95% CI 1.06 to 1.46; 1 study, 115 participants; low‐certainty evidence), but may result in little to no difference in rates of recurrence (5.4% versus 7.8%, RR 0.69, 95% CI 0.30 to 1.61; 6 studies, 551 participants; very low‐certainty evidence), wound infection (2.8% versus 6.4%, RR 0.44, 95% CI 0.16 to 1.17; 6 studies, 559 participants; very low‐certainty evidence), and wound dehiscence (2.5% versus 3.0%, RR 0.82, 95% CI 0.17 to 3.84; 3 studies, 250 participants; very low‐certainty evidence) compared to tension‐free midline closure. Furthermore, off‐midline closure may result in longer time to return to work compared to tension‐free midline closure (MD 3.00 days, 95% CI 1.52 to 4.48 days; 1 study, 60 participants; low‐certainty evidence). There were no data available for quality of life.

Karydakis flap versus Limberg flap

Karydakis flap probably results in little to no difference in time to wound healing compared to Limberg flap (MD 0.36 days, 95% CI ‐1.49 to 2.22; 6 studies, 526 participants; moderate‐certainty evidence). Compared to Limberg flap, Karydakis flap may result in little to no difference in the proportion of wounds healed (80.0% versus 66.7%, RR 1.20, 95% CI 0.77 to 1.86; 1 study, 30 participants; low‐certainty evidence), recurrence rate (5.1% versus 4.5%, RR 1.14, 95% CI 0.61 to 2.14; 9 studies, 890 participants; low‐certainty evidence), wound infection (7.9% versus 5.1%, RR 1.55, 95% CI 0.90 to 2.68; 8 studies, 869 participants; low‐certainty evidence), wound dehiscence (7.4% versus 6.2%, RR 1.20, 95% CI 0.41 to 3.50; 7 studies, 776 participants; low‐certainty evidence), and time to return to work (MD ‐0.23 days, 95% CI ‐5.53 to 5.08 days; 6 studies, 541 participants; low‐certainty evidence). There were no data available for quality of life.

Authors' conclusions

This Cochrane review examines the midline and off‐midline wound closure options for pilonidal sinus, predominantly based on young adult studies. Off‐midline flap procedures demonstrate there may be benefits over conventional midline closure for pilonidal sinus, with various off‐midline flap techniques. When off‐midline flap closures were compared to tension‐free midline closure, low‐certainty evidence indicated there may be improved wound healing and increased time to return to work for off‐midline closure, whilst very low‐certainty evidence indicated there may be no evidence of a difference in other outcomes. There may be no evidence of an advantage found amongst the off‐midline techniques evaluated. The choice of either procedure is likely to be based on a clinician's preference, experience, patient characteristics, and the patients' preferences. To more accurately determine the benefits and potential harms of these closure techniques, further large‐scale and meticulously‐designed trials are essential. Specifically, there is a pressing need for more studies addressing the paediatric population, in addition to adult studies.

Plain language summary

Midline or off‐midline wound closure techniques: which works better to treat pilonidal sinus disease?

Key messages

• Closing wounds from one side of the butt crack (off‐midline closure) may lead to faster healing, lower chances of the disease returning, fewer infections, less chance of the wound breaking open again, shorter hospital stays, quicker return to work, and less pain compared to closing wounds directly from the middle of the butt crack (conventional midline closure).

• We could not find enough good‐quality evidence to show that off‐midline closure is better than closing wounds from the middle of the butt crack using special methods to reduce tension on the wound edges (tension‐free midline closure).

• Larger, well‐designed studies are needed to give better estimates of the benefits and potential harms of the various midline and off‐midline wound closure techniques for pilonidal sinus in both children and adults.

What is pilonidal sinus, and how can it be treated?

Pilonidal sinus disease is a condition where hairs get trapped in a tiny tunnel in the skin, usually near the top of the butt crack. It mostly affects young adults and can cause different symptoms, from no symptoms to significant pain. Surgery to remove the tunnel and surrounding skin is the main treatment for long‐lasting (chronic) pilonidal sinus.

Closing the surgical wound from the middle of the butt crack is called midline closure. In conventional midline closure, the wound is directly closed in the middle of the butt crack. In tension‐free midline closure, special methods are used to reduce tension on the wound edges.

Closing the wound from one side of the butt crack is called off‐midline closure. There are many methods for off‐midline closure, including various flap methods. These use a flap of skin to cover the area where tissue has been removed. Therefore, in this review, unless otherwise specified, the flaps mentioned are one of the off‐midline closure methods.

What did we want to find out?

We wanted to find out:

• how well do the different methods of closing wounds work for treating pilonidal sinus?

• how well do different off‐midline methods of closing wounds work for treating pilonidal sinus?

We wanted to find out how different methods of closing wounds affect:

• the time for wounds to heal completely;

• the proportion of people whose wounds heal completely;

• the proportion of pilonidal sinus recurrence;

• the proportion of wound infection;

• the proportion of people who had a wound breaking open again (wound dehiscence);

• the time to return to work;

• quality of life after surgery.

What did we do?

We searched for studies that investigated off‐midline closure compared to conventional or tension‐free midline closure, and studies that compared different types of off‐midline closure. We compared and summarised the results of the studies and rated our confidence in the evidence, based on factors such as study methods and sizes.

What did we find?

We found 33 studies that involved a total of 3667 people. The median or average age of the people in the studies ranged from 21 to 34 years. Most were men. The studies were primarily conducted in the Middle East. We identified 9 comparisons. In this summary, we present the results of the 3 main comparisons:

Off‐midline closure versus conventional midline closure

• Off‐midline closure probably causes a large reduction in time for wounds to heal completely, reduces the recurrence of pilonidal sinus, and reduces wound infections.

• Off‐midline closure may cause a large reduction in wound dehiscence and a quicker return to work.

• Off‐midline closure may have little to no effect on the proportion of people whose wounds heal completely, but we are very uncertain about the results.

Off‐midline closure versus tension‐free midline closure

• Off‐midline closure may cause a large reduction in time for wounds to heal completely, may increase the proportion of people whose wounds heal completely at 3 months, and may increase the time to return to work.

• Off‐midline closure may have little to no effect on the recurrence of pilonidal sinus, wound infections, and wound dehiscence.

Karydakis flap versus Limberg flap

• The Karydakis flap probably has little to no effect on time for wounds to heal completely, the recurrence of pilonidal sinus, wound infections, wound dehiscence, and time to return to work when compared to the Limberg flap.

What are the limitations of the evidence?

Our confidence in the findings was limited by poor study design, a small number of cases in the included studies, not enough studies to draw definitive conclusions, and a lack of data on certain outcomes in some studies.

The results of further research could differ from some results of this review.

How up to date is this evidence?

The evidence is current to June 2022.

Summary of findings

Summary of findings 1. Summary of findings table ‐ Off‐midline closure compared to conventional midline closure for pilonidal sinus.

Off‐midline closure compared to conventional midline closure for pilonidal sinus
Patient or population: pilonidal sinus Setting: hospital Intervention: off‐midline closure Comparison: conventional midline closure
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with conventional midline closure	Risk with off‐midline closure
Time to wound healing	The mean time to wound healing ranged from 15.3 to 23 days	MD 5.23 days lower (7.55 lower to 2.92 lower)	‐	300 (3 RCTs)	⊕⊕⊕⊝ Moderatea
Proportion of wounds healed follow‐up: range 9 days to 11 days	885 per 1000	1000 per 1000 (814 to 1000)	RR 1.13 (0.92 to 1.39)	207 (2 RCTs)	⊕⊝⊝⊝ Very lowb,c
Recurrence rate follow‐up: range 3 months to 36 months	68 per 1000	15 per 1000 (8 to 31)	RR 0.22 (0.11 to 0.45)	1492 (13 RCTs)	⊕⊕⊕⊝ Moderated	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Wound infection	117 per 1000	38 per 1000 (26 to 58)	RR 0.32 (0.22 to 0.49)	1568 (13 RCTs)	⊕⊕⊕⊝ Moderated	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Wound dehiscence	89 per 1000	39 per 1000 (24 to 63)	RR 0.44 (0.27 to 0.71)	1389 (11 RCTs)	⊕⊕⊝⊝ Lowd,e
Time to return to work	The mean time to return to work ranged from 8.68 to 21.60 days	MD 3.72 days lower (6.11 lower to 1.33 lower)	‐	820 (6 RCTs)	⊕⊕⊝⊝ Lowa,f
Quality of life ‐ not reported	‐	‐	‐	‐	‐
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440579146797591733.

^a We downgraded by one level for indirectness due to due to the inclusion of only two techniques (Limberg flap and modified Limberg flap) in the analysis, while several other off‐midline closure techniques for pilonidal sinus exist.
^b We downgraded by one level for indirectness due to the inclusion of only two techniques (Limberg flap and Karydakis flap) in the analysis, while several other off‐midline closure techniques for pilonidal sinus exist.
^c We downgraded by two levels for imprecision because of a small number of events and a wide CI.
^d We downgraded by one level for imprecision because the number of events is small.
^e We downgraded by one level for publication bias due to funnel plot asymmetry.
^f We downgraded by one level for inconsistency due to considerable true heterogeneity in effect estimates. While the majority of differences are between small and large treatment effects favoring the intervention, one study showed an effect in the opposite direction.

Summary of findings 2. Summary of findings table ‐ Off‐midline closure compared to tension‐free midline closure for pilonidal sinus.

Off‐midline closure compared to tension‐free midline closure for pilonidal sinus
Patient or population: pilonidal sinus Setting: hospital Intervention: off‐midline closure Comparison: tension‐free midline closure
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with tension‐free midline closure	Risk with off‐midline closure
Time to wound healing	The median time to wound healing was 51 days	median 14 days higher (12 higher to 20 higher)	‐	116 (1 RCT)	⊕⊕⊝⊝ Lowa,b	This study only described the release of tension using lateral dissection (Hemmingsson 2022).
Proportion of wounds healed follow‐up: 3 months	764 per 1000	947 per 1000 (809 to 1000)	RR 1.24 (1.06 to 1.46)	115 (1 RCT)	⊕⊕⊝⊝ Lowa,c	This study only described the release of tension using lateral dissection (Hemmingsson 2022).
Recurrence rate follow‐up: range 11 days to 167 months	78 per 1000	54 per 1000 (23 to 125)	RR 0.69 (0.30 to 1.61)	551 (6 RCTs)	⊕⊝⊝⊝ Very lowd,e	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results. Only two studies delineated the specifics of the tension‐free method (Okuş 2012; Sevinc 2016), noting the release of subcutaneous tissue on both sides of the incision line across spans of 2 cm to 5 cm.
Wound infection	64 per 1000	28 per 1000 (10 to 75)	RR 0.44 (0.16 to 1.17)	559 (6 RCTs)	⊕⊝⊝⊝ Very lowd,e	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results. Only two studies delineated the specifics of the tension‐free method (Okuş 2012; Sevinc 2016), noting the release of subcutaneous tissue on both sides of the incision line across spans of 2 cm to 5 cm.
Wound dehiscence	30 per 1000	25 per 1000 (5 to 115)	RR 0.82 (0.17 to 3.84)	250 (3 RCTs)	⊕⊝⊝⊝ Very lowe,f	Only one study delineated the specifics of the tension‐free method (Sevinc 2016), noting the release of subcutaneous tissue on both sides of the incision line across spans of 3 cm to 5 cm.
Time to return to work	The mean time to return to work was 20 days	MD 3 days higher (1.52 higher to 4.48 higher)	‐	60 (1 RCT)	⊕⊕⊝⊝ Lowb,f	This study only described the release of tension using lateral dissection (Arnous 2019).
Quality of life ‐ not reported	‐	‐	‐	‐	‐
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_444316868680523960.

^a We downgraded by one level for indirectness due to the inclusion of only one technique (Karydakis flap) in the analysis, while several other off‐midline closure techniques for pilonidal sinus exist.
^b We downgraded by one level for imprecision because the number of events is small.
^c We downgraded by one level for imprecision due to CI boundary closest to no effect.
^d We downgraded by one level for indirectness due to the inclusion of only two techniques (Karydakis flap and Limberg flap) in the analysis, while several other off‐midline closure techniques for pilonidal sinus exist.
^e We downgraded by two levels for imprecision because of a small number of events and a wide CI.
^f We downgraded by one level for indirectness due to the inclusion of only one technique (Limberg flap) in the analysis, while several other off‐midline closure techniques for pilonidal sinus exist.

Summary of findings 3. Summary of findings table ‐ Off‐midline closure (Karydakis flap) compared to off‐midline closure (Limberg flap) for pilonidal sinus.

Off‐midline closure (Karydakis flap) compared to off‐midline closure (Limberg flap) for pilonidal sinus
Patient or population: pilonidal sinus Setting: hospital Intervention: off‐midline closure (Karydakis flap) Comparison: off‐midline closure (Limberg flap)
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	№ of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with off‐midline closure (Limberg flap)	Risk with off‐midline closure (Karydakis flap)
Time to wound healing	The mean time to wound healing ranged from 9.56 to 32.6 days	MD 0.36 days more (1.49 fewer to 2.22 more)	‐	526 (6 RCTs)	⊕⊕⊕⊝ Moderatea	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Proportion of wounds healed	667 per 1000	800 per 1000 (513 to 1000)	RR 1.20 (0.77 to 1.86)	30 (1 RCT)	⊕⊕⊝⊝ Lowb	Follow‐up was not reported.
Recurrence rate follow‐up: range 6 months to 41 months	45 per 1000	51 per 1000 (28 to 97)	RR 1.14 (0.61 to 2.14)	890 (9 RCTs)	⊕⊕⊝⊝ Lowc	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Wound infection	51 per 1000	79 per 1000 (46 to 136)	RR 1.55 (0.90 to 2.68)	869 (8 RCTs)	⊕⊕⊝⊝ Lowc	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Wound dehiscence	62 per 1000	74 per 1000 (25 to 217)	RR 1.20 (0.41 to 3.50)	776 (7 RCTs)	⊕⊕⊝⊝ Lowc	Although there was a minority of studies with a high risk of bias, our sensitivity analysis revealed that the inclusion of these high‐risk studies did not significantly impact the overall pooled results.
Time to return to work	The mean time to return to work ranged from 8.21 to 28.66 days	MD 0.23 days fewer (5.53 fewer to 5.08 more)	‐	541 (6 RCTs)	⊕⊕⊝⊝ Lowa,d
Quality of life ‐ not reported	‐	‐	‐	‐	‐
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; MD: mean difference; RR: risk ratio
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
See interactive version of this table: https://gdt.gradepro.org/presentations/#/isof/isof_question_revman_web_440579156710304952.

^a We downgraded by one level for inconsistency due to substantial or considerable heterogeneity in effect estimates. 
^b We downgraded by two levels for imprecision due to the small sample size and number of events and a wide CI.
^c We downgraded by two levels for imprecision because of a small number of events and a wide CI.
^d We downgraded by one level for imprecision due to the wide CI. Although the minimal clinically important difference has not been established, the clinical context should be considered when interpreting these results.

Background

See Appendix 1 for a glossary of terms.

Description of the condition

Pilonidal sinus disease is a prevalent and debilitating condition characterised by the presence of hair follicles in the upper part of the natal cleft of the buttocks (Iesalnieks 2019; Johnson 2021). The term 'pilonidal sinus' was first introduced by Hodges in 1880 to describe a chronic aberrant tract between the buttocks, derived from the Latin words 'pilus' (hair) and 'nidus' (nest) (Hodges 1880). During World War II, the disease was colloquially termed 'Jeep disease' due to its high prevalence amongst jeep‐driving soldiers (Buie 1944). Presently, the terms 'pilonidal sinus disease', 'pilonidal disease', and 'pilonidal sinus' are frequently used interchangeably.

The prevalence of pilonidal sinus has been steadily increasing (Iesalnieks 2019; Ommer 2021). In Germany, its incidence amongst males rose from 43 to 56 cases per 100,000 people per year between 2005 and 2017, while the incidence amongst females increased from 14 to 18 cases per 100,000 people per year within the same period (Oetzmann 2021). In the USA, approximately 70,000 new cases are reported annually (Johnson 2019). A meta‐analysis of studies from different regions showed that women constitute around 21% of the patients with pilonidal sinus disease (Luedi 2021). This disease primarily affects young adults: for example, one study showed that the average age of presentation was 20.8 years (ranging from 17 to 28 years) in Turkey, with men and women presenting at the average age of 19.3 and 21.6 years, respectively (Duman 2017).

The exact aetiology of pilonidal sinus remains unclear, but it is considered an acquired disease linked to hair and inflammation in the sacrococcygeal area (Johnson 2019). Patey and Scarff proposed that hair in the natal cleft punctures the skin, embedding in the subcutaneous tissues and triggering a foreign‐body reaction that leads to abscesses or chronic sinus formation (Patey 1946). Based on this hypothesis, Karydakis identified three key risk factors for pilonidal disease, including the mechanical properties of hair, factors that increase the force of hair insertion, and skin vulnerability factors (Karydakis 1992).

The clinical presentation of pilonidal sinus disease varies widely, from asymptomatic cases to acute infections or chronic exacerbations characterised by inflammation and drainage (Johnson 2021). Asymptomatic forms are marked by one or more non‐inflamed pits in the natal cleft, usually discovered incidentally (Iesalnieks 2019). Acute pilonidal abscesses manifest as swelling and pain in the natal cleft, resulting from infected hair follicles and pus accumulation in the fat tissue. In cases of chronic disease, continuous or intermittent serous or purulent discharge may occur, either from the pits (rare) or from lateral secondary openings (Iesalnieks 2019).

Description of the intervention

The ideal surgical treatment for pilonidal sinus should aim to reduce disease recurrence and limit the extent of excision. It should also facilitate a swift return to daily activities for patients, minimise work absenteeism, and leave a subtle scar (Isik 2016).

The cornerstone of acute pilonidal abscess treatment is incision and drainage (ABCMDA 2023; Herrod 2022; Iesalnieks 2021; Johnson 2019; Milone 2021). For managing chronic pilonidal sinus, validated surgical approaches include minimally invasive methods, healing through primary closure, excision with marsupialisation, and healing by secondary intention (when a wound is left open, rather than being stitched together, and left to heal by itself) (Johnson 2019; Milone 2021).

Although minimally invasive approaches represent a treatment option for chronic pilonidal disease, data in the current literature are still limited (Milone 2021). Primary closure involves wound closure either immediately after surgery or following a delay, positioned on the natal cleft's midline or off‐midline (including flap techniques). Amongst various flap techniques, the Karydakis flap (Karydakis 1973), the Bascom cleft lift (Bascom 2007), and the Limberg flap (Azab 1984) are the most widely used and recommended by several guidelines (ABCMDA 2023; Iesalnieks 2021; Johnson 2019; Milone 2021). Other documented flap techniques include V‐Y flap (Schoeller 1997) and Z‐plasty techniques (Hodgson 1981). Secondary intention healing involves leaving the surgical wound open to heal naturally from the base upwards. A previous Cochrane review concluded that both secondary intention healing and primary closure have their pros and cons, with patient treatment goals also guiding the surgical approach selection (Al‐Khamis 2010). A survey found that patients, when given a choice, would opt for procedures with a low risk of infection/wound complications and a high chance of cure (Wickramasekera 2023). However, they would trade these attributes for a quicker recovery, and there are age‐related differences in preferences (Wickramasekera 2023).

Postoperative care principles focus on infection prevention, promoting healing from the wound base, and achieving rapid re‐epithelialisation with optimal cosmetic outcomes (Timmons 2007). In general, postoperative care for pilonidal sinus, including patient positioning, peri‐wound cleaning, wound cleaning, and patient education, should be tailored to the wound assessment and the patient's preferences (Harris 2016).

How the intervention might work

Midline closure is a surgical technique that positions the postoperative wound in the midline. However, due to the high wound morbidity and recurrence rates associated with midline closure (Al‐Khamis 2010), three major guidelines advise against its use (Iesalnieks 2021; Johnson 2019; Milone 2021). Nevertheless, recent advancements in tension‐free midline closure techniques, which involve a subcutaneous release at the level of the fascia of the gluteus maximus muscle to flatten the wound after midline closure, appear to address these concerns (Okuş 2012; Sevinç 2016; Youssef 2015).

Off‐midline closure procedures include lateral incision and flap techniques. For superficial and simple sinuses, the removal of the pit or sinus through a lateral incision is generally adequate (Soll 2011). However, the sinus tract of pilonidal sinus is often complex, necessitating extensive resection that leaves a sizeable defect difficult to suture directly (Romaniszyn 2020). To resolve this issue, surgeons frequently employ flap techniques, excising the diseased area and replacing it with healthy tissue to cover and repair the defect resulting from extensive resection.

Why it is important to do this review

A 2010 Cochrane review comparing secondary and primary closure healing after surgical treatment for pilonidal sinus emphasised the advantages of off‐midline closure over midline closure (Al‐Khamis 2010). Since its publication, several randomised controlled trials (RCTs) have provided further evidence regarding the clinical effectiveness of midline and off‐midline wound closure methods. Additionally, there is a need to assess the role of tension‐free midline closure in managing pilonidal sinus disease (Okuş 2012; Sevinç 2016; Youssef 2015). The optimal primary closure method remains unclear. Consequently, it is crucial to evaluate systematically the effects of midline and off‐midline wound closure methods after surgical treatment for pilonidal sinus in order to inform clinical practice (Salandra 2022). In this Cochrane review, we aim to address the following two questions.

• What are the effects of off‐midline wound closure in comparison to conventional or tension‐free midline wound closure following surgical treatment for pilonidal sinus?

• Which off‐midline flap closure is the most effective?

We hope the results of this review will assist surgeons in making evidence‐based decisions.

Objectives

To assess the effects of midline and off‐midline wound closure methods for pilonidal sinus, and to determine the optimal off‐midline flap procedures.

Methods

Criteria for considering studies for this review

Types of studies

We included RCTs with parallel designs, either at the cluster or individual level. We excluded cross‐over trials, as the nature of our research question did not allow for cross‐over designs. We also excluded quasi‐randomised trials, in which the method for allocating participants to different treatments was not strictly random (e.g. by date of birth, day of the week, month of the year, medical record number, or order of inclusion in the study).

We included studies reported as full text, those published as abstracts only, and unpublished data. There was no language restriction.

Types of participants

We included participants of any age and sex undergoing midline and off‐midline wound closure methods after surgical treatment for pilonidal sinus disease. No distinction was made between new and recurrent presentations of pilonidal sinus. We excluded participants presenting with an abscess, as incision and drainage represent the standard intervention for acute abscesses in pilonidal sinus disease, according to current guidelines (ABCMDA 2023; Iesalnieks 2021; Johnson 2019; Milone 2021). In studies with mixed populations (i.e. studies in which some participants met the inclusion criteria and others did not, such as studies whose sample included some participants with an abscess), we included only the eligible participants, if this information was reported separately, or if we were able to obtain it from study authors. Otherwise, we planned to include studies with mixed populations where more than 80% of the participants met the eligibility criteria.

Types of interventions

For this review, we considered midline closure (including conventional procedures and tension‐free procedures without flap techniques) and all types of off‐midline closure (e.g. Karydakis flap, modified Karydakis flap, cleft lift procedure, Limberg flap, modified Limberg flap).

We included the following comparisons:

• off‐midline closure versus conventional midline closure;

• off‐midline closure versus tension‐free midline closure;

• one type of off‐midline flap closure versus another type of off‐midline flap closure.

Types of outcome measures

We analysed the following outcomes in the review, but did not use them as a basis for including or excluding studies.

Primary outcomes

• Wound healing, presented in either or both of the following formats

  • Time to wound healing (time in days until wound healed), assessed clinically by researchers or the clinical team. Where possible, we aimed to present time‐to‐event data as hazard ratios (HRs) with 95% confidence intervals (CIs). Mean time‐to‐healing data were only included where we were certain that all wounds had healed.

  • Proportion of wounds healed (number of wounds healed/not healed) at short‐ (30‐day), medium‐ (6‐month), and long‐term (1‐year) follow‐up, assessed clinically by researchers or a clinical team

• Recurrence rate (proportion of treated incidences of pilonidal sinus that recurred during the longest follow‐up reported in the study), assessed clinically by researchers or the clinical team

Secondary outcomes

• Adverse effects (wound infection, wound dehiscence): proportion of participants in each group with an adverse effect during study follow‐up, assessed clinically by researchers or the clinical team

• Time to return to work (days), as described during the follow‐up time in the study. Where possible, we aimed to present data as time‐to‐event.

• Operative time (minutes)

• Length of hospital stay (days)

• Postoperative or post‐procedural pain on the first postoperative day. We included pain only where studies reported mean scores with a standard deviation using a validated scale, such as a visual analogue scale (VAS).

• Quality of life (QoL) measured by validated questionnaires, such as the Medical Outcomes Study 36‐item short‐form health survey (SF‐36; Ware 1992), EuroQol 5‐dimension questionnaire (EQ‐5D; EuroQol Group 1990), or Cardiff Wound Impact Schedule (CWIS; Price 2004), during study follow‐up

• Cost‐effectiveness: within‐study cost‐effectiveness analysis comparing mean differences in effects with mean cost differences between the two arms. We planned to extract data on incremental mean cost per incremental gain in benefit (incremental cost‐effectiveness ratio (ICER)). We also considered other measures of relative cost‐effectiveness (e.g. net monetary benefit, net health benefit).

Timing of outcome assessment

When multiple time points were reported for the outcomes, we included the longest reported follow‐up period in the main analyses. We also considered the following time points for the primary and secondary outcomes:

• less than 12 months of follow‐up;

• 12 months of follow‐up or more.

Hierarchy of outcome assessment

If there was more than one scale for an outcome, we opted for the validated one. If all the scales were validated, we selected the one that was provided more frequently across all the included studies, in order to reduce heterogeneity.

Search methods for identification of studies

Electronic searches

We searched the following electronic databases to identify reports of relevant clinical trials:

• Cochrane Wounds Specialised Register (searched 28 June 2022);

• Cochrane Central Register of Controlled Trials (CENTRAL; 2022, Issue 5) in the Cochrane Library (searched 28 June 2022);

• MEDLINE Ovid including In‐Process & Other Non‐Indexed Citations (1946 to 28 June 2022);

• Embase Ovid (1974 to 28 June 2022);

• CINAHL Plus EBSCO (Cumulative Index to Nursing and Allied Health Literature) (1937 to 28 June 2022).

The search strategies for the Cochrane Wounds Specialised Register, CENTRAL, MEDLINE Ovid, Embase Ovid and CINAHL Plus EBSCO can be found in Appendix 2. In MEDLINE Ovid, we combined the subject‐specific strategy with the sensitivity‐ and precision‐maximising version of the Cochrane Highly Sensitive Search Strategy for identifying randomised trials (2008 revision) (Lefebvre 2022). We combined the Embase Ovid search with the Ovid Embase filter developed by Cochrane UK (Lefebvre 2022). We combined the CINAHL Plus EBSCO search with the trial filter developed by Glanville 2019. There were no restrictions with respect to language, date of publication, or study setting.

We also searched the following clinical trials registries:

• US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov; searched 28 June 2022);

• World Health Organization International Clinical Trials Registry Platform (www.who.int/clinical-trials-registry-platform; searched 28 June 2022).

Search strategies for clinical trial registries can be found in Appendix 2.

Searching other resources

We aimed to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, as well as relevant systematic reviews, meta‐analyses, and health technology assessment reports.

We contacted authors of key papers and abstracts to request further information about their trials.

We did not perform a separate search for adverse effects of interventions used; we considered the adverse effects described in the included studies only.

Data collection and analysis

Selection of studies

We used the reference management software EndNote 20 to merge search results and remove duplicate records (EndNote 2013), so that each study, rather than each report, was the unit of interest in the review. We then imported these results into the latest version of Covidence for screening and filtering of titles and abstracts. Two review authors (Cai Z, Zhao Z) independently screened the titles and abstracts to identify potentially eligible studies. Two review authors (Cai Z, Zhao Z) then independently screened the full texts of the selected records to identify studies for inclusion, recording reasons for the exclusion of the ineligible studies. We resolved any disagreements through discussion amongst all review authors. Two review authors (Cai Z, Zhao Z) recorded the selection process in sufficient detail to complete a Preferred Reporting Items for Systematic Reviews and Meta‐Analyses (PRISMA) flow diagram (Page 2021), and a Characteristics of excluded studies table. Abstracts and full texts in all languages were considered for inclusion. For non‐English language papers, we used Google Translate or Baidu Translate in the first instance to assist with eligibility assessment. If needed, we planned to seek translators in our university networks or through Cochrane Engage to assist with assessing the eligibility of studies and to assist with data extraction in any studies deemed eligible.

Data extraction and management

We used a data collection form for study characteristics and outcome data. We piloted the data extraction form to assess its usability. Two review authors (Cai Z, Zhao Z) independently extracted study characteristics from the included studies using Microsoft Excel (Microsoft 2016). A third review author (Zhang B) spot‐checked study characteristics for accuracy against the study reports. We extracted the following study characteristics.

• Methods: study design, sample size calculation, randomisation, blinding, CONSORT flow diagram

• Participants: location, study setting, number, baseline characteristics (including age, sex, body mass index, type of pilonidal sinus (e.g. primary, recurrent, or not specified)), inclusion criteria, and exclusion criteria

• Interventions: intervention and comparison (details of operation performed and any co‐interventions), duration of follow‐up

• Outcomes: primary and secondary outcomes specified and collected (we extracted the study authors’ definition and the measure used to assess each reported outcome), and time points reported

• Publication status of study

• Notes: funding for study, and notable declarations of interest of study authors

Two review authors (Cai Z, Zhao Z) independently extracted outcome data from the included studies. We extracted the number of events and number of participants per treatment group for dichotomous outcomes, and means and standard deviations (SDs) and number of participants per treatment group for continuous outcomes. We noted in the Characteristics of included studies table if outcome data were not reported in a usable way and when data were transformed or estimated from a graph. We resolved disagreements through discussion amongst all review authors. One review author (Cai Z) transferred data into the latest version of Review Manager Web (RevMan Web 2024). We double‐checked that data were entered correctly by comparing the data presented in the systematic review with those in the study reports. If multiple time points existed, we selected the longest follow‐up from each study.

We planned to use WebPlotDigitizer or Engauge Digitizer to extract data from graphs or figures (Mitchell 2019; Rohatgi 2021). We intended for these data to be extracted in duplicate. However, we did not find any data that needed to be extracted from graphs or figures.

Assessment of risk of bias in included studies

Two review authors (Zhao Z, Cai Z) independently assessed the risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022b) and the risk of bias (RoB) 2 tool (Sterne 2019; Appendix 3). In the context of this review, we were interested in quantifying the effect of assignment to the interventions at baseline, regardless of whether the interventions were received as intended (i.e. the intention‐to‐treat (ITT) effect). We assessed the risk of bias for the seven outcomes presented in each of the three summary of findings tables for the main comparisons (see: Table 1; Table 2; Table 3). The risk of bias assessments fed into the risk of bias domain of the GRADE approach for assessing the certainty of a body of evidence. Furthermore, we also assessed the risk of bias for these seven outcomes for the remaining six comparisons. We resolved any disagreement through discussion amongst all review authors. We assessed the risk of bias according to the following domains:

• bias arising from the randomisation process;

• bias due to deviations from intended interventions;

• bias due to missing outcome data;

• bias in measurement of the outcome;

• bias in selection of the reported result.

We judged each domain to be at ‘low risk of bias’, ‘some concerns’, or ‘high risk of bias'. Where the risk of bias was unclear due to lack of information or uncertainty of potential for bias, we contacted study authors to seek clarification. We used the data available in the published report if we did not receive a response within four weeks.

We also presented an overall risk of bias judgement for each study by evaluating the risk of bias across all five domains. We considered the risk of bias for each outcome as follows:

• low risk of bias: if all domains for this outcome were deemed to be at low risk of bias.

• some concerns: there were some concerns for at least one domain for this result, but none of the domains were judged to be at high risk of bias.

• high risk of bias: at least one domain was judged to be at high risk of bias, or there were some concerns for multiple domains in a way that substantially lowered confidence in the result.

We used the RoB 2 Excel tool to implement RoB 2 (Sterne 2019) and stored RoB 2 data as supplementary files. We used RevMan Web 2024 to input the risk of bias results and visualise RoB 2 data in the following ways:

• risk of bias traffic lights alongside the forest plots in the analyses;

• RoB tables showing RoB 2 domain judgements and support for judgements.

For cluster‐randomised clinical trials, we planned to consider an additional domain that specifically applies to the design of the cluster‐randomised clinical trial: 'Bias arising from the timing of identification and recruitment of individual participants within clusters in relation to timing of randomisation' (RoB 2 Domain 1b). We planned to follow the suggested algorithm for reaching risk of bias judgements for bias arising from the timing of identification and recruitment of participants in a cluster‐randomised trial (Eldridge 2021; Higgins 2022c; Appendix 4). At the time of review preparation, we planned to use the most recent recommendations for assessing risk of bias in cluster‐randomised trials. However, we did not identify any cluster‐randomised clinical trials.

Assessment of bias in conducting the systematic review

We conducted the review according to this published protocol and reported any deviation from it in the 'Differences between protocol and review' section of this Cochrane review.

Measures of treatment effect

We analysed dichotomous data as risk ratios (RRs) (Deeks 2022), with 95% CIs. If available, we planned to extract and report hazard ratios (HRs) with 95% CIs for time‐to‐event data. We analysed continuous data as mean difference (MD) or standardised mean difference (SMD), depending on whether the same scale was used to measure an outcome, with 95% CIs. We entered data presented as a scale with a consistent direction of effect across studies.

When the included studies used different scales to measure the same conceptual outcome (e.g. quality of life), we planned to calculate SMDs with the corresponding 95% CIs. We planned to back‐translate SMDs to a typical scale by multiplying the SMD by a typical among‐person standard deviation (Schünemann 2022b). However, we did not identify any instances where different scales were used to measure the same conceptual outcome.

Unit of analysis issues

Where multiple study arms were reported in a single study, we only included the relevant arms. If a study had more than two eligible intervention arms or control arms, we described all relevant study conditions in the qualitative data synthesis and, where possible, combined groups to create a single pair‐wise comparison, as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022c).

We planned to follow the guidance on cluster‐randomised trials in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022c), but we did not find any cluster‐randomised trials to include in this review.

Dealing with missing data

We adhered to the five general recommendations for addressing missing data outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022). We contacted study authors to confirm key study characteristics and obtain any missing numerical outcome data whenever possible, such as when we identified a study as an abstract only or when data were unavailable for all participants. Where this was not possible, and the missing data were thought to introduce serious bias, we explored the impact of including such studies on the overall assessment of results through a sensitivity analysis. We clearly described any assumptions or imputations used to handle missing data, and we planned to explore the effect of imputation by conducting sensitivity analyses.

For dichotomous outcomes (e.g. recurrence), we analysed only the available data (i.e. ignoring the missing data) and assessed the risk of bias due to missing outcome data in randomised trials.

For continuous outcomes (e.g. change in QoL), we calculated the MD or SMD based on the number of participants analysed at that time point. If the number of participants analysed was not presented for each time point, we used the number of randomised participants in each group at baseline.

When possible, we computed missing SDs from other statistics, such as standard errors, CIs, or P values, following the methods recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2022a). In cases where sample means and SDs could not be calculated, we estimated these figures using the five‐number summary (Luo 2018; Wan 2014). In the absence of available mean values, we utilised median values as a substitute. If neither calculation nor estimation was feasible, we documented this and excluded the study from any relevant meta‐analyses conducted.

Assessment of heterogeneity

Assessment of heterogeneity can be a complex, multifaceted process. We considered clinical and methodological heterogeneity (i.e. variation in study participants, interventions, outcomes, and characteristics, such as length of follow‐up). These assessments of clinical and methodological heterogeneity were supplemented, where appropriate, by information regarding statistical heterogeneity, assessed using the Chi² test in conjunction with the I² measure (Higgins 2003).

Where data were available, and it was appropriate, we used forest plots to evaluate the direction and magnitude of the effects and the degree of overlap between CIs. We used the I² measure to assess heterogeneity amongst the studies in each analysis (Higgins 2003). If we identified substantial heterogeneity, we reported it and investigated possible causes by following the recommendations in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022).

We followed the rough guide to the interpretation of I² values in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022), which states:

• 0% to 40%: might not be important;

• 30% to 60%: may represent moderate heterogeneity*;

• 50% to 90%: may represent substantial heterogeneity*;

• 75% to 100%: considerable heterogeneity*.

*As noted in the Cochrane Handbook (Deeks 2022), we kept in mind that the importance of I² depends on: (i) magnitude and direction of effects and (ii) strength of evidence for heterogeneity (e.g. P value from Chi² test, or a confidence interval for I²; uncertainty in the value of the I² measure is substantial when the number of studies is small). The Chi² test was interpreted such that a P value of 0.10 or less indicated evidence of statistical heterogeneity.

Assessment of reporting biases

We created and examined a funnel plot to explore possible small study biases for outcomes with more than 10 studies (Page 2022). We used Egger's test to determine the statistical significance of reporting bias (Egger 1997), and considered a P value of less than 0.05 to be a statistically significant reporting bias.

Data synthesis

We used RevMan Web 2024 for data synthesis. We conducted an initial qualitative assessment of all the included studies to examine whether pooling of results (meta‐analysis) was reasonable. This took into account differences in study populations, inclusion and exclusion criteria, interventions, and outcome assessment. Where clinical and methodological heterogeneity was thought to be acceptable, we considered conducting meta‐analyses. We meta‐analysed even when statistical heterogeneity was high, but attempted to interpret the causes for this heterogeneity using subgroup analyses and meta‐regression for that purpose, if possible.

We pooled outcomes grouped by comparison. We used a random‐effects model by default because a certain degree of heterogeneity was expected amongst studies. We assumed that studies were not all estimating the same intervention effect and that such intervention effects followed a normal distribution across studies (DerSimonian 1986).

If we were unable to pool data statistically using meta‐analysis, we planned to conduct a narrative synthesis of results, adhering to the Synthesis Without Meta‐analysis (SWiM) guidelines (Campbell 2020).

Subgroup analysis and investigation of heterogeneity

Considering that effect modifiers influence how well interventions work (Deeks 2022), we planned to carry out the following subgroup analyses for the outcomes included in the summary of findings tables.

• Types of off‐midline closure procedure (e.g. Karydakis flap, Limberg flap, and others) (default subgroup for comparisons 1 and 2 only, with no numerical limits)

• Disease status (e.g. new, recurrent, or not specified), as recurrent disease continues to be a problem for patients and surgeons alike

• Perioperative antibiotic prophylaxis (e.g. yes, no, or not specified), as co‐interventions may play a role in intervention effects

We extracted data at the longest follow‐up time point from each study. Due to a lack of consistency across studies, we conducted a subgroup analysis for follow‐up duration.

Although we initially planned to perform a subgroup analysis to explore differences in disease status and perioperative antibiotic prophylaxis, studies frequently lacked clarity in reporting data on clinical subgroups. In the comparison between the Karydakis and Limberg flaps, two studies included recurrent disease cases. In the remaining comparisons, studies had unspecified disease status or involved primary disease cases.

We restricted subgroup analyses to outcomes where at least three studies were available. We used the formal test for subgroup interactions in RevMan Web 2024 and exercised caution in the interpretation of subgroup analyses, as advised in the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2022).

Sensitivity analysis

If we were able to include at least three studies for the outcomes presented in the summary of findings tables, we performed a sensitivity analysis to determine whether the conclusions were robust to decisions made during the review process.

• We repeated the analysis excluding studies at an overall high risk of bias for the outcome of interest.

• We repeated the analysis excluding studies that did not provide clear tension‐free method descriptions for the comparison of off‐midline closure versus tension‐free midline closure (added post‐protocol).

Interpreting results and reaching conclusions

We followed the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions for interpreting results (Schünemann 2022b), being careful to distinguish a lack of evidence of effect from a lack of effect. We based our conclusions only on findings from the quantitative or narrative synthesis of included studies for this review. We avoided making recommendations for practice. Our Implications for research suggest priorities for future research and outline what the remaining uncertainties are in the area.

Summary of findings and assessment of the certainty of the evidence

We created summary of findings tables, according to the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2022a). We created summary of findings tables for the following comparisons and critical outcomes.

• Comparisons

  • Off‐midline closure versus conventional midline closure

  • Off‐midline closure versus tension‐free midline closure

  • Off‐midline (Karydakis flap) versus off‐midline (Limberg flap) closure

• Outcomes

  • Time to wound healing

  • Proportion of wounds healed (during short‐term follow‐up)

  • Recurrence rate

  • Wound infection

  • Wound dehiscence

  • Time to return to work

  • Quality of life

In the summary of findings tables, we prioritised the longest time point available for outcomes. Two review authors (Cai Z, Zhao Z) independently assessed the certainty of the evidence. We used the overall RoB 2 judgement to feed into the GRADE assessment. We used the five GRADE considerations (study limitations, consistency of effect, imprecision, indirectness, and publication bias) to assess the certainty of a body of evidence as it relates to the studies that contributed data to the meta‐analyses for the prespecified outcomes, and we reported the certainty of evidence as high, moderate, low, or very low. We used methods and recommendations described in the Cochrane Handbook for Systematic Reviews of Interventions (Schünemann 2022a; Schünemann 2022b). We used the latest version of GRADEpro GDT to prepare the summary of findings tables. We justified all decisions to downgrade the certainty of evidence using footnotes, and we made comments to aid the readers' understanding of the review where necessary.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification; Characteristics of ongoing studies.

Results of the search

We identified 905 records from the initial search of the prespecified databases, and four additional references from handsearching the reference lists of potentially relevant studies. After removing duplicates, a total of 795 records remained for title and abstract screening. We discarded 666 records as irrelevant to the review, and retrieved the remaining 129 records in full text for closer inspection. Working independently, two review authors (Cai Z, Zhao Z) identified 33 studies (with a total of 41 references) that met the review's inclusion criteria. We excluded 64 studies (with 67 references) as ineligible, and we have provided details of nine of these excluded studies in the Characteristics of excluded studies table (see Excluded studies). We listed 13 studies (14 references) as 'awaiting classification', and identified six ongoing studies (seven references). A study flow diagram is presented in Figure 1.

1.

Study flow diagram

Included studies

See Characteristics of included studies and Table 4.

1. Description of included studies.

Study ID	Country/Region	Group	Intervention	Male (n, %)	Age, median (range), years	Follow‐up period, median (range), months	Sample randomised	Sample analysed
Off‐midline versus midline (conventional)
Abu 1999	United Arab Emirates	Total	‐	37, 80.4%	NR	18.0*	46	46
		Experimental	Limberg flap	19, 79.2%	23.0* (17.0 to 30.0)	18.0*	24	24
		Comparator	Midline closure	18, 81.8%	21.0* (18.0 to 34.0)	18.0*	22	22
Akca 2005	Turkey	Total	‐	168, 84.0%	NR	28.0 (23.0 to 38.0)	200	200
		Experimental	Limberg flap	85, 85.0%	26.0* (15.0 to 60.0)	28.0 (23.0 to 36.0)	100	100
		Comparator	Midline closure	83, 83.0%	28.0* (17.0 to 43.0)	29.0 (23.0 to 38.0)	100	100
Bukhari 2020	Pakistan	Total	‐	69, 56.6%	NR	NR	122	122
		Experimental	Limberg flap	38, 62.3%	28.2 ± 6.2*	NR	61	61
		Comparator	Midline closure	31, 50.8%	27.5 ± 5.8*	NR	61	61
Daglar 2004	Turkey	Total	‐	58, 84.1%	25.0* (16.0 to 35.0)	NR	69	69
		Experimental	Limberg flap	NR	NR	31.5* (13.0 to 52.0)	37	37
		Comparator	Midline closure	NR	NR	28.4* (12.0 to 89.0)	32	32
Dass 2012	India	Total	‐	74, 92.5%	28.4 ± 8.6* (16.0 to 48.0)	36.0‡	80	80
		Experimental	Limberg flap	38, 95.0%	NR	36.0‡	40	40
		Comparator	Midline closure	36, 90.0%	NR	36.0‡	40	40
Ertan 2005	Turkey	Total	‐	91, 91.0%	NR	19.0*	100	100
		Experimental	Limberg flap	46, 92.0%	28.0 ± 7.7*	NR	50	50
		Comparator	Midline closure	45, 90.0%	26.3 ± 6.8*	NR	50	50
Khan 2013	India	Total	‐	104, 86.7%	NR	24.0‡	120	120
		Experimental	Limberg flap	53, 88.3%	24.0* (17.0 to 42.0)	NR	60	60
		Comparator	Midline closure	51, 85.0%	26.0* (16.0 to 40.0)	NR	60	60
Maghsudi 2020	Iran	Total	‐	62, 62.0%	NR (14.0 to 36.0)	3.0*	100	100
		Experimental	Limberg flap	29, 58.0%	24.0 ± 4.6* (14.0 to 36.0)	3.0*	50	50
		Comparator	Midline closure	33, 66.0%	24.1 ± 4.5* (14.0 to 34.0)	3.0*	50	50
Muzi 2010	Italy	Total	‐	223, 86.0%	NR	NR	272	260
		Experimental	Limberg flap	114, 88.0%	25.1 ± 7.5*	45.74 ± 7.6*	135	130
		Comparator	Midline closure	109, 84.0%	25.0 ± 8.6*	47.83 ± 6.6*	137	130
Galal 2012	Egypt	Total	‐	91, 75.8%	NR	NR	120	120
		Experimental	Modified Limberg flap	45, 75.0%	26.0 ± 8.6*	22.0 ± 7.64*	60	60
		Comparator	Midline closure	46, 76.7%	27.0 ± 9.4*	21.5 ± 6.82*	60	60
Shabbir 2014	Pakistan	Total	‐	55, 91.7%	NR	12.0‡	60	60
		Experimental	Modified Limberg flap	27, 90.0%	NR	12.0‡	30	30
		Comparator	Midline closure	28, 93.3%	NR	12.0‡	30	30
Sakr 2006	Egypt	Total	‐	138, 85.7%	NR (14.0 to 43.0)	30.4* (12.0 to 49.0)	170	161
		Experimental	Karydakis flap	68, 86.1%	26.5* (18.0 to 43.0)	NR	NR	79
		Comparator	Midline closure	70, 85.4%	22.7* (14.0 to 37.0)	NR	NR	82
Enshaei 2014	Iran	Total	‐	51, 63.8%	24.2± 6.1* (15.0 to 40.0)	6.0*	80	80
		Experimental	Rotational Flap	29, 72.5%	NR	6.0*	40	40
		Comparator	Midline closure	22, 55.0%	NR	6.0*	40	40
Enshaei 2020	Iran	Total	‐	84, 87.5%	26.6 ± 5.0*	6.0*	100	96
		Experimental	Rotational Flap	41, 87.2%	27.3 ± 4.5*	6.0*	50	47
		Comparator	Midline closure	43, 87.8%	25.8 ± 5.5*	6.0*	50	49
Off‐midline versus midline (tension‐free)
Arnous 2019	Egypt	Total	‐	52, 86.6%	24.1*	NR	62	60
		Experimental	Limberg flap	27, 90.0%	23.8 ± 5.9*	16.4 ± 3.7*	31	30
		Comparator	Tension‐free primary closure	25, 83.3%	25.1 ± 7.4*	17.0 ± 4.0*	31	30
Hemmingsson 2022	Sweden	Total	‐	103, 88.8%	NR	129.5 (0.4 to 166.7)	125	116
		Experimental	Karydakis flap	52, 86.7%	26.5 (95% CI 23.0 to 30.0)	NR	64	60
		Comparator	Tension‐free primary closure	51, 91.1%	28.0 (95% CI 23.0 to 29.0)	NR	61	56
Okuş 2012	Turkey	Total	‐	88, 94.6%	25.0 (17.0 to 43.0)	29.5 (8.0 to 43.0)	100	93
		Experimental	Limberg flap	48, 98.0%	24.0 (17.0 to 33.0)	31.5 (8.0 to 42.0)	NR	49
		Comparator	Tension‐free primary closure	40, 90.9%	25.5 (17.0 to 43.0)	28.0 (8.0 to 43.0)	NR	44
Rashidian 2014	Iran	Total	‐	NR	27.6§* (14.0 to 49.0)	18.0§* (3.0 to 36.0)	40	40
		Experimental	Limberg flap	NR	NR	NR	20	20
		Comparator	Tension‐free primary closure	NR	NR	NR	20	20
Sevinc 2016	Turkey	Total	‐	123, 82.0%	24.0 (18.0 to 47.0)	24.2 (18.5 to 34.3)	150	150
		Experimental 1	Limberg flap	39, 78.0%	23.5 (18.0 to 46.0)	NR	50	50
		Experimental 2	Karydakis flap	41, 82.0%	24.0 (18.0 to 38.0)	NR	50	50
		Comparator	Tension‐free primary closure	43, 86.0%	23.5 (18.0 to 47.0)	NR	50	50
Tavassoli 2011	Iran	Total	‐	80, 80.0%	24.0 ± 4.35* (16.0 to 36.0)	6.0*	100	100
		Experimental	Limberg flap	NR	NR	6.0*	50	50
		Comparator	Tension‐free primary closure	NR	NR	6.0*	50	50
Off‐midline (Karydakis flap) versus off‐midline (Limberg flap)
Alvandipour 2019	Iran	Total	‐	30, 46.9%	NR	6.0*	65	64
		Experimental	Karydakis flap	21, 56.8%	25.9 ± 9.5*	6.0*	37	37
		Comparator	Limberg flap	9, 33.3%	34.2 ± 10.5*	6.0*	28	27
Anwar 2021	Pakistan	Total	‐	62, 91.1%	NR	6.0*	70	69
		Experimental	Karydakis flap	32, 94.1%	26.5 ± 6.0*	6.0*	35	35
		Comparator	Limberg flap	30, 88.2%	26.0 ± 6.0*	6.0*	35	34
Arslan 2014**	Turkey	Total	‐	172, 86.4%	24.0 (18.0 to 39.0)	33.0 (24.0 to 41.0)	330	295
		Experimental 1	Modified Limberg flap	95, 88.0%	24.7 ± 5.1*	32.9 ± 4.6*	110	108
		Experimental 2	Karydakis flap	77, 84.6%	24.7 ± 5.1*	33.3 ± 5.4*	110	91
		Comparator	Limberg flap	68, 70.8%	26.5 ± 5.9*	34.5 ± 5.3*	110	96
Bali 2015	Turkey	Total	‐	64, 90.1%	NR	28.0*	71	71
		Experimental	Karydakis flap	32, 94.1%	23.5*	27.6 ± 8.3*	34	34
		Comparator	Limberg flap	32, 86.5%	25.0*	28.2 ± 7.7*	37	37
El Hadidi 2019	Egypt	Total	‐	115, 95.8%	22.0 (18.0 to 40.0)	23.0* (18.0 to 28.0)	120	120
		Experimental	Karydakis flap	57, 95%	22.0 (18.0 to 40.0)	NR	60	60
		Comparator	Limberg flap	58, 96.7%	22.0 (19.0 to 35.0)	NR	60	60
Elhiny 2021	Egypt	Total	‐	23, 42.6%	32.0 ± 7.9* (19.0 to 48.0)	12.0*	54	54
		Experimental	Karydakis flap	12, 44.4%	33.0 ± 6.9* (21.0 to 45.0)	12.0*	27	27
		Comparator	Limberg flap	11, 40.7%	29.8 ± 8.4* (19.0 to 48.0)	12.0*	27	27
Kohla 2016	Egypt	Total	‐	NR	NR (18.0 to 32.0)	NR	30	30
		Experimental	Karydakis flap	NR	NR (18.0 to 25.0)	NR	15	15
		Comparator	Limberg flap	NR	NR (19.0 to 32.0)	NR	15	15
Riaz 2019	Pakistan	Total	‐	166, 81.3%	27.2 ± 8.2*	6.0*	204	204
		Experimental	Karydakis flap	80, 78.4%	26.4 ± 7.7*	NR	102	102
		Comparator	Limberg flap	86, 84.3%	28.1 ± 8.7*	NR	102	102
Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap)
Guner 2013	Turkey	Total	‐	98, 80.3%	25.0* (18.0 to 48.0)	13.0* (6.0 to 18.0)	122	122
		Experimental	Bascom’s cleft lift	50, 82.0%	24.6* (18.0 to 46.0)	13.0* (6.0 to 18.0)	61	61
		Comparator	Limberg flap	48, 78.7%	25.4* (18.0 to 48.0)	13.0* (6.0 to 18.0)	61	61
Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap)
Can 2010	Turkey	Total	‐	143, 98.6%	22.0 (19.0 to 40.0)	16.8 ± 3.4*	145	145
		Experimental	Modified Limberg flap	76, 98.7%	22.0 (19.0 to 40.0)	NR	72	77
		Comparator	Karydakis flap	67, 98.5%	22.0 (20.0 to 31.0)	NR	73	68
Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap)
Kartal 2019	Turkey	Total	‐	83, 83.0%	NR	25.0* (18.0 to 33.0)	100	100
		Experimental	Modified Karydakis flap	42, 84.0%	24.3 ± 7.4*	NR	50	50
		Comparator	Karydakis flap	41, 82.0%	24.8 ± 7.0*	NR	50	50
Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap)
Bessa 2013	Turkey	Total	‐	112, 93.3%	23.0 (15.0 to 41.0)	20.5 (6.0 to 36.0)	125	120
		Experimental	Modified Limberg flap	58, 96.7%	23.0 (16.0 to 41.0)	NR	62	60
		Comparator	Modified Karydakis flap	54, 90.0%	23.0 (15.0 to 36.0)	NR	63	60
Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap)
Saydam 2015	Turkey	Total	‐	100, 100%	23.0 (19.0 to 34.0)	12.0*	100	100
		Experimental	Modified Limberg	50, 100%	23.0 (20.0 to 34.0)	NR	50	50
		Comparator	Burow’s triangle advancement flap	50, 100%	23.0 (19.0 to 31.0)	NR	50	50

CI: confidence interval; NR: not reported; SD: standard deviation

^‡Follow‐up periods are indicated as either "followed up for a period of" or "over a follow‐up of".

^§The presented data also include results from a 'lay open' surgical group.

*Data are presented as means, or as mean ± SD.

**Arslan 2014 contributed data for three comparisons:

1. Off‐midline (Karydakis flap) versus off‐midline (Limberg flap)
2. Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap) techniques
3. Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap)

Design

All included studies were RCTs. Of these, 11 studies either adhered to the CONSORT statement or included a CONSORT flow diagram in their reports (Schulz 2010). These studies are: Alvandipour 2019; Anwar 2021; Arnous 2019; Bali 2015; Bessa 2013; Can 2010; El Hadidi 2019; Guner 2013; Hemmingsson 2022; Muzi 2010; and Sevinc 2016.

Setting

All trials were carried out in a hospital setting. Of these, three were day‐case surgeries, as documented in Bessa 2013, Hemmingsson 2022, and Kartal 2019. In Muzi 2010, only primary closures were day‐case procedures.

One study was published in Turkish (Daglar 2004), while the remaining RCTs were in English.

Geographical distribution of studies:

• Turkey: 12 studies (Akca 2005; Arslan 2014; Bali 2015; Bessa 2013; Can 2010; Daglar 2004; Ertan 2005; Guner 2013; Kartal 2019; Okuş 2012; Saydam 2015; Sevinc 2016)

• Egypt: six studies (Arnous 2019; El Hadidi 2019; Elhiny 2021; Galal 2012; Kohla 2016; Sakr 2006)

• Iran: six studies (Alvandipour 2019; Enshaei 2014; Enshaei 2020; Maghsudi 2020; Rashidian 2014; Tavassoli 2011)

• Pakistan: four studies (Anwar 2021; Bukhari 2020; Riaz 2019; Shabbir 2014)

• India: two studies (Dass 2012; Khan 2013)

• Italy: one study (Muzi 2010)

• Sweden: one study (Hemmingsson 2022)

• United Arab Emirates: one study (Abu 1999)

Participants

We included 33 studies, comprising 3752 randomised participants, of which 3667 were analysed. The median or mean age of participants in all included studies was between 21.0 and 34.2 years. Two studies included participants with recurrent disease (Bali 2015; El Hadidi 2019), while the remaining studies had unspecified disease status (Akca 2005; Arnous 2019; Elhiny 2021; Enshaei 2014; Hemmingsson 2022; Khan 2013; Kohla 2016; Muzi 2010; Okuş 2012), or involved primary disease. Summary details of the participants at baseline are given in Table 4.

Types of interventions

Wound closure methods

We included studies assessing a wide variety of wound closure methods.

• Midline closure

  • Conventional midline closure: this category encompasses all midline closures where the skin and subcutaneous tissue are not released.

  • Tension‐free midline closure: this category encompasses all midline closures where the skin and subcutaneous tissue are released (Arnous 2019; Hemmingsson 2022; Okuş 2012; Rashidian 2014; Sevinc 2016; Tavassoli 2011).

    • Four studies simply described the release of tension using lateral dissection or by creating "small flaps with the gluteus maximus muscles on either side" (Arnous 2019; Hemmingsson 2022; Rashidian 2014; Tavassoli 2011).

    • Only two studies delineated the specifics of the tension‐free method (Okuş 2012; Sevinc 2016), noting the release of subcutaneous tissue on both sides of the incision line across spans of 2 cm to 5 cm.

    • To understand the potential impact of these details, we conducted a sensitivity analysis excluding studies that did not provide clear procedural descriptions.

• Off‐midline closure

  • Bascom’s cleft lift

  • Limberg flap

  • Karydakis flap

  • Modified Limberg flap

  • Modified Karydakis flap

  • Burow’s triangle advancement flap

Comparisons

• Off‐midline closure versus conventional midline closure (Abu 1999; Akca 2005; Bukhari 2020; Daglar 2004; Dass 2012; Enshaei 2014; Enshaei 2020; Ertan 2005; Galal 2012; Khan 2013; Maghsudi 2020; Muzi 2010; Sakr 2006; Shabbir 2014)

• Off‐midline closure versus tension‐free midline closure (Arnous 2019; Hemmingsson 2022; Okuş 2012; Rashidian 2014; Sevinc 2016; Tavassoli 2011)

• Off‐midline (Karydakis flap) versus off‐midline (Limberg flap) (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Kohla 2016; Riaz 2019; Sevinc 2016)

• Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap) (Guner 2013)

• Off‐midline (Modified Limberg flap) versus off‐midline (Limberg flap) (Arslan 2014)

• Off‐midline (Modified Limberg flap) versus off‐midline (Karydakis flap) (Arslan 2014; Can 2010)

• Off‐midline (Modified Karydakis flap) versus off‐midline (Karydakis flap) (Kartal 2019)

• Off‐midline (Modified Limberg flap) versus off‐midline (Modified Karydakis flap) (Bessa 2013)

• Off‐midline (Modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap) (Saydam 2015)

Excluded studies

We excluded 64 studies (67 references) at the full‐text screening stage because they did not meet the Criteria for considering studies for this review. Forty‐three studies did not have the correct intervention or comparison; 20 studies did not qualify as RCTs (within this group, we identified seven quasi‐randomised trials); and one study was withdrawn.

To provide more in‐depth insights, we describe nine excluded studies in the Characteristics of excluded studies table (Ates 2011; Awad 2007; Cihan 2006; Ersoy 2009; Karim 2020; Kayal 2014; Nordon 2009; Tokac 2015; Zorlu 2016). We highlighted these studies as readers of the review might perceive them as both relevant and potentially eligible.

Studies awaiting classification

We listed 13 studies as 'awaiting classification'. Five studies are listed as completed in trial registries, but their results have not yet been published (IRCT20180719040525N1; NCT01268969; NCT03765086; NCT04017260; NCT04572919). Additionally, we listed three meeting abstracts with insufficient publicly‐available information as awaiting classification (Bayhan 2020; Khalil 2019; Wright 2001). For further details, please refer to the Characteristics of studies awaiting classification section.

Ahmad 2018, which compared off‐midline closure to conventional midline closure, reported an unusually older demographic. Typically, pilonidal sinus affects mainly teenagers and young adults (Duman 2017; Oetzmann 2021). However, in Ahmad 2018, the experimental group had a mean age of 41.6 years, while the control group's mean age was reported to be 34.8 years. Additionally, the study presented an unexpected significant difference in baseline characteristics between intervention groups. Unfortunately, we were unable to obtain further details from the study authors regarding their baseline characteristics.

The following four studies reported information that was ambiguous, conflicting, or raised questions for us (Kaleem 2021; Khan 2016; Sabet 2016; Youssef 2015).

• In Kaleem 2021, which compared off‐midline closure to conventional midline closure, the experimental and control groups reported mean ages of 44.37 and 45.24 years, respectively. The method section of Kaleem 2021 refers to this "descriptive cross‐sectional study", and further states that "367 sample size was calculated" and "patients were randomly divided into two equal groups using the lottery method". The study design, as presented, was puzzling.

• Khan 2016, which compared the modified Limberg flap to the modified Karydakis flap, also reported an unusually older demographic. In this study, the modified Limberg flap group had a mean age of 42.71 ± 14.78 years, while the modified Karydakis flap group had a mean age of 43.42 ± 16.66 years.

• Sabet 2016, which contrasted off‐midline closure with tension‐free midline closure, indicated operation durations for both the Limberg flap group (16.1 ± 3.4 minutes) and the tension‐free midline closure group (14.9 ± 3.4 minutes). These durations seem unusually brief in the context of our existing knowledge.

• Youssef 2015, another study comparing off‐midline closure to tension‐free midline closure, presented a disparity in its report. The study indicated that by the third postoperative month, the tension‐free group had significantly less bodily pain and enhanced social functioning compared to the modified Limberg flap group. Yet, our inspection of the data suggested that the tension‐free group had higher (i.e. worse) pain scores and lower social functioning scores than the modified Limberg flap group. Additionally, the Mental Component Summary score was higher for the tension‐free group compared to the flap group. Due to these discrepancies, we've approached the authors to provide clarity on the SF‐36 data.

We have reached out to the authors of these studies to seek further clarification. Until we receive responses that clarify the issues and questions highlighted, we will withhold analyses of these studies.

Ongoing studies

We identified six ongoing studies (IRCT201710109014N195; NCT03914729; NCT04217824; NCT05289297; NL4763; NL4792). For further details, please refer to the Characteristics of ongoing studies table.

Risk of bias in included studies

For each outcome we prespecified for risk of bias assessment, the results‐level RoB 2 tables include the judgements and support for judgements for each domain and the overall risk of bias (Table 55; Table 56; Table 57; Table 58; Table 59; Table 60; Table 61; Table 62; Table 63; Table 64; Table 66; Table 67; Table 68; Table 69; Table 70; Table 71; Table 72; Table 73; Table 74; Table 75; Table 76; Table 77; Table 78; Table 79; Table 80; Table 81; Table 82; Table 83; Table 84; Table 85; Table 86; Table 87; Table 88; Table 89; Table 90; Table 91; Table 92). Domain‐specific judgements and supporting statements for each outcome in each study can be found at the following link (https://osf.io/download/qvnj6/).

Risk of bias for analysis 1.1 Time to wound healing.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.1.1 Limberg flap

Enshaei 2014

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. According to the chi‐square test, there is no significant difference between sexes in two surgical procedures (P=0.1). The mean age of the patients was 24.17± 6.05 minimum 15 and maximum 40.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. Measurement of the outcome may be same between intervention groups. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Maghsudi 2020

Low risk of bias

A block randomization method (including qua druple blocks) was used for this purpose. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 24.0 ± 4.6 (range 14–36); control group: 24.1 ± 4.5 (range 14–34); Gender: Intervention group: 29(58%) male, 21(42%) female; control group: 33(66%) male, 17(34%) female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

All patients returned to the clinic after 1 week, 1 month, and 3 months following discharge and were evaluated for recurrence, wound dehiscence, surgical site infection, edema of the flap, seroma, mobilization, time of return to work, complete healing, and ability to sit on the toilet without pain. "The person evaluating the outcome will be blinded to the allocated group". Observer‐reported outcomes involving some judgement.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at en.irct.ir (IRCT20180606039985N1). "Patients will be evaluated for recurrence and complications and improvement rate after one week, one months and three months post surgery". To compare quantitative data between groups, the one‐way analysis of variance test was used.

Some concerns

Based on judgements of previous domains.

Subgroup 1.1.2 Modified Limberg flap

Galal 2012

Low risk of bias

Randomization was achieved through a computer‐generated schedule and the results were sealed into envelopes. The envelopes were drawn and opened by a nurse in the operating room. There was no significant difference between both groups regarding age, sex, preoperative symptoms, and period of follow‐up.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. For quantitative data Student’s t‐test was used to compare between two groups.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 1.2 Proportion of wounds healed.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.2.1 Limberg flap

Abu 1999

Low risk of bias

Forty six consecutive patients were randomly allocated by a system of sequentially numbered sealed envelopes. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. Evaluation of healing includes qualitative and quantitative analysis of chronic ulcers and wounds. "Patients were assessed in the outpatient clinic and admitted under a single team." Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.2.2 Karydakis flap

Sakr 2006

Low risk of bias

Written consent was obtained from all patients who were categorized, using the closed envelope method, into 2 groups. “All baseline differences were not statistically significant”.

Low risk of bias

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 1.3 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.3.1 Limberg flap

Abu 1999

Low risk of bias

Forty six consecutive patients were randomly allocated by a system of sequentially numbered sealed envelopes. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Akca 2005

Low risk of bias

Randomisation method: A computer‐generated table of random numbers. The randomisation was carried out by an independent computer consultant. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Daglar 2004

Some concerns

The patients were randomly divided into two groups. There is no information available regarding the randomization process and allocation concealment in the study. The two groups were homogenized in terms of age, gender, occupation, body mass index (preoperative inflammation or presence of infection, preoperative antibiotic use, preoperative length of stay.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Dass 2012

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: mean 28.4±8.6 years (range 16–48 years). Gender: Intervention group: 38 male, 2 female. Control group: 36 male, 4 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "For qualitative data, Chisquare test was used to see the significant difference in proportion between two groups".

Some concerns

Based on judgements of previous domains.

Ertan 2005

Low risk of bias

Allocation was performed by balanced randomization after informed consent had been obtained. There was no significant difference between the groups in age, sex, duration of symptoms, and preoperative VAS scores.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Khan 2013

Low risk of bias

After obtaining written informed consent, patients were randomly (using table of random numbers) allocated to undergo either excision and primary closure (group A, n=60) or rhomboid excision and the Limberg flap procedure (group B, n=60). There were no significant differences between the two groups with respect to age and sex (P>0.05).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The chi‐square test was applied for comparison between frequencies".

Some concerns

Based on judgements of previous domains.

Maghsudi 2020

Low risk of bias

A block randomization method (including qua druple blocks) was used for this purpose. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 24.0 ± 4.6 (range 14–36); control group: 24.1 ± 4.5 (range 14–34); Gender: Intervention group: 29(58%) male, 21(42%) female; control group: 33(66%) male, 17(34%) female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"Patients will be evaluated for recurrence and complications and improvement rate after one week, one months and three months post surgery. Recurrence outcome at 3 month follow‐up“. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at en.irct.ir (IRCT20180606039985N1). The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Recurrence outcome at 3 month follow‐up". "To compare qualitative data, the chi‐square test was used".

Some concerns

Based on judgements of previous domains.

Muzi 2010

Low risk of bias

Eligible patients were assigned randomly to 1 of the 2 surgical groups according to a computer‐generated list. Allocations were sealed in opaque numbered envelopes. The code was broken on the day of surgery before the induction of anesthetics. There was no significant difference between the groups regarding sex, age, and presence of acute infection.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Some concerns

"Lost in follow‐up: 7/130 vs 5/130, Recurrence rate: 5/130 vs 0/130". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 7/130 vs 5/130, Recurrence rate: 5/130 vs 0/130". Extreme case: 12/130 vs 0/130, P=0.0004; 5/130 vs 5/130, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Lost in follow‐up: 7/130 vs 5/130, P=0.769; Recurrence rate: 5/130 vs 0/130, P=0.061". "During the follow‐up period, 7 patients, who previously had undergone primary closure, were lost to follow‐up evaluation, of these 2 with normal findings were not motivated to continue their check‐ups. In the Limberg flap arm, 5 patients were lost at follow‐up evaluation, of these, 1 with normal findings was not motivated to continue their check‐ups". No difference between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Some concerns

"Sub sequently, the patients were followed up with telephone interviews." "An independent observer, who was not from the surgical team and who was unaware of the treatment assignments, recorded all data, which included postoperative events and follow‐up findings". Outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The end points were categoric, either binary (yes/no) or ordinal, and have been tested using the Fisher exact test for the binary end points and the Kruskal–Wallis test for the ordinal end points".

Some concerns

Based on judgements of previous domains.

Subgroup 1.3.2 Modified Limberg flap

Galal 2012

Low risk of bias

Randomization was achieved through a computer‐generated schedule and the results were sealed into envelopes. The envelopes were drawn and opened by a nurse in the operating room. There was no significant difference between both groups regarding age, sex, preoperative symptoms, and period of follow‐up.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Chi square test, and Fisher’s exact test was used for qualitative data".

Some concerns

Based on judgements of previous domains.

Shabbir 2014

Low risk of bias

A total of 60 patients were randomly divided into two equal groups using the lottery method to prevent selection bias. Each group had 30(50%) patients. In group 1 there were 27(90%) males and 3(10%) females, while group 2 had 28(93.3%) males and 2(6.7%) females.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Both groups were followed up for twelve months (after one week, two weeks, six months and twelve months) for recurrence".

Some concerns

Based on judgements of previous domains.

Subgroup 1.3.3 Karydakis flap

Sakr 2006

Low risk of bias

Written consent was obtained from all patients who were categorized, using the closed envelope method, into 2 groups. “All baseline differences were not statistically significant”.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.3.4 Rotational flap

Enshaei 2014

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. According to the chi‐square test, there is no significant difference between sexes in two surgical procedures (P=0.1). The mean age of the patients was 24.17± 6.05 minimum 15 and maximum 40.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

No information provided. Measurement of the outcome may be same between intervention groups. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Enshaei 2020

Low risk of bias

For randomization, a number of sequences were assigned by an independent researcher using random allocation software version 1 in permuted blocks size of 3 and 6. At the time of randomization, a researcher who was not involved in the study process opened a numbered envelope containing patient allocation sheets in one of the three study groups, and the participants were thus assigned to any of the three study groups. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: mean 27.32 ± 4.5 years; Control group: mean 25.84 ± 5.5 years. Gender: Intervention group: 41 male, 6 female; Control group: 43 male, 6 female. BMI: Intervention group: 26.8 ± 1.9; Control group: 27.3 ± 2.2.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"Following the discharge, we asked the patients to visit the outpatient clinic at the end of the first, second and forth weeks, as well as 3 and 6 months after surgery. In each visit, the checklist was completed from the day after surgery up to 6 months in terms of recurrence, surgical complications, post‐operative pain, and duration of hospitalization". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "For comparison of the rates, we used chi‐square tests."

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 1.5 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.5.1 Limberg flap

Akca 2005

Low risk of bias

Randomisation method: A computer‐generated table of random numbers. The randomisation was carried out by an independent computer consultant. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Bukhari 2020

Low risk of bias

"Applying lottery method, patients were randomly allocated into two groups." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: mean 28.23±6.15 years; Control group: mean 27.49±5.79 years. Gender: Intervention group: 38 male, 23 female; Control group: 31 male, 30 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "By applying the Chi‐Square test, the difference between two groups was found to be statistically significant in terms of wound infection (p>0.014), edema (p>0.008) and wound dehiscence (p>0.017)."

Some concerns

Based on judgements of previous domains.

Daglar 2004

Some concerns

The patients were randomly divided into two groups. There is no information available regarding the randomization process and allocation concealment in the study. The two groups were homogenized in terms of age, gender, occupation, body mass index (preoperative inflammation or presence of infection, preoperative antibiotic use, preoperative length of stay.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Dass 2012

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: mean 28.4±8.6 years (range 16–48 years). Gender: Intervention group: 38 male, 2 female. Control group: 36 male, 4 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "For qualitative data, Chisquare test was used to see the significant difference in proportion between two groups".

Some concerns

Based on judgements of previous domains.

Ertan 2005

Low risk of bias

Allocation was performed by balanced randomization after informed consent had been obtained. There was no significant difference between the groups in age, sex, duration of symptoms, and preoperative VAS scores.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Statistical analysis was performed using Chi‐square or Fisher’s Exact test to compare discrete variables".

Some concerns

Based on judgements of previous domains.

Khan 2013

Low risk of bias

After obtaining written informed consent, patients were randomly (using table of random numbers) allocated to undergo either excision and primary closure (group A, n=60) or rhomboid excision and the Limberg flap procedure (group B, n=60). There were no significant differences between the two groups with respect to age and sex (P>0.05).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The chi‐square test was applied for comparison between frequencies".

Some concerns

Based on judgements of previous domains.

Maghsudi 2020

Low risk of bias

A block randomization method (including qua druple blocks) was used for this purpose. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 24.0 ± 4.6 (range 14–36); control group: 24.1 ± 4.5 (range 14–34); Gender: Intervention group: 29(58%) male, 21(42%) female; control group: 33(66%) male, 17(34%) female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at en.irct.ir (IRCT20180606039985N1). The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Muzi 2010

Low risk of bias

Eligible patients were assigned randomly to 1 of the 2 surgical groups according to a computer‐generated list. Allocations were sealed in opaque numbered envelopes. The code was broken on the day of surgery before the induction of anesthetics. There was no significant difference between the groups regarding sex, age, and presence of acute infection.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

High risk of bias

"Lost in follow‐up: 7/130 vs 5/130, wound infection: 14/130 vs 4/130". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 7/130 vs 5/130, wound infection: 14/130 vs 4/130". Extreme case: 21/130 vs 4/130, P=0.0005; 14/130 vs 9/130, P=0.383. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Lost in follow‐up: 7/130 vs 5/130, P=0.769; wound infection: 14/130 vs 4/130, P=0.025". "During the follow‐up period, 7 patients, who previously had undergone primary closure, were lost to follow‐up evaluation, of these 2 with normal findings were not motivated to continue their check‐ups. In the Limberg flap arm, 5 patients were lost at follow‐up evaluation, of these, 1 with normal findings was not motivated to continue their check‐ups". No difference between intervention groups in the proportions of missing outcome data. Differences between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

Noninfection was defined as pilonidal disease in absence of pain, acute inflammation, and chronic inflammation. An independent observer, who was not from the surgical team and who was unaware of the treatment assignments, recorded all data, which included postoperative events and follow‐up findings.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Subgroup 1.5.2 Modified Limberg flap

Galal 2012

Low risk of bias

Randomization was achieved through a computer‐generated schedule and the results were sealed into envelopes. The envelopes were drawn and opened by a nurse in the operating room. There was no significant difference between both groups regarding age, sex, preoperative symptoms, and period of follow‐up.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"Infection was considered as leakage of purulent secretion through the surgical wound and not only peri‐incisional hyperaemia". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data. "Chi square test, and Fisher’s exact test was used for qualitative data".

Some concerns

Based on judgements of previous domains.

Shabbir 2014

Low risk of bias

A total of 60 patients were randomly divided into two equal groups using the lottery method to prevent selection bias. Each group had 30(50%) patients. In group 1 there were 27(90%) males and 3(10%) females, while group 2 had 28(93.3%) males and 2(6.7%) females.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.5.3 Karydakis flap

Sakr 2006

Low risk of bias

Written consent was obtained from all patients who were categorized, using the closed envelope method, into 2 groups. “All baseline differences were not statistically significant”.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.5.4 Rotational flap

Enshaei 2014

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. According to the chi‐square test, there is no significant difference between sexes in two surgical procedures (P=0.1). The mean age of the patients was 24.17± 6.05 minimum 15 and maximum 40.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"Depending on the clinical status, patients discharged and the first and fourth weeks and 6 months postoperatively were visited and questionnaires were completed". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Then, depending on the clinical status, patients discharged and the first and fourth weeks and 6 months postoperatively were visited and questionnaires were completed."

Some concerns

Based on judgements of previous domains.

Enshaei 2020

Low risk of bias

For randomization, a number of sequences were assigned by an independent researcher using random allocation software version 1 in permuted blocks size of 3 and 6. At the time of randomization, a researcher who was not involved in the study process opened a numbered envelope containing patient allocation sheets in one of the three study groups, and the participants were thus assigned to any of the three study groups. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: mean 27.32 ± 4.5 years; Control group: mean 25.84 ± 5.5 years. Gender: Intervention group: 41 male, 6 female; Control group: 43 male, 6 female. BMI: Intervention group: 26.8 ± 1.9; Control group: 27.3 ± 2.2.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "For comparison of the rates, we used chi‐square tests."

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 1.7 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.7.1 Limberg flap

Abu 1999

Low risk of bias

Forty six consecutive patients were randomly allocated by a system of sequentially numbered sealed envelopes. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Akca 2005

Low risk of bias

Randomisation method: A computer‐generated table of random numbers. The randomisation was carried out by an independent computer consultant. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Bukhari 2020

Low risk of bias

"Applying lottery method, patients were randomly allocated into two groups." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: mean 28.23±6.15 years; Control group: mean 27.49±5.79 years. Gender: Intervention group: 38 male, 23 female; Control group: 31 male, 30 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements.

Some concerns

Based on judgements of previous domains.

Dass 2012

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: mean 28.4±8.6 years (range 16–48 years). Gender: Intervention group: 38 male, 2 female. Control group: 36 male, 4 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "For qualitative data, Chisquare test was used to see the significant difference in proportion between two groups".

Some concerns

Based on judgements of previous domains.

Ertan 2005

Low risk of bias

Allocation was performed by balanced randomization after informed consent had been obtained. There was no significant difference between the groups in age, sex, duration of symptoms, and preoperative VAS scores.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Statistical analysis was performed using Chi‐square or Fisher’s Exact test to compare discrete variables".

Some concerns

Based on judgements of previous domains.

Khan 2013

Low risk of bias

After obtaining written informed consent, patients were randomly (using table of random numbers) allocated to undergo either excision and primary closure (group A, n=60) or rhomboid excision and the Limberg flap procedure (group B, n=60). There were no significant differences between the two groups with respect to age and sex (P>0.05).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The chi‐square test was applied for comparison between frequencies".

Some concerns

Based on judgements of previous domains.

Maghsudi 2020

Low risk of bias

A block randomization method (including qua druple blocks) was used for this purpose. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 24.0 ± 4.6 (range 14–36); control group: 24.1 ± 4.5 (range 14–34); Gender: Intervention group: 29(58%) male, 21(42%) female; control group: 33(66%) male, 17(34%) female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at en.irct.ir (IRCT20180606039985N1). The results were unlikely to have been selected from multiple eligible outcome measurements.

Some concerns

Based on judgements of previous domains.

Muzi 2010

Low risk of bias

Eligible patients were assigned randomly to 1 of the 2 surgical groups according to a computer‐generated list. Allocations were sealed in opaque numbered envelopes. The code was broken on the day of surgery before the induction of anesthetics. There was no significant difference between the groups regarding sex, age, and presence of acute infection.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Some concerns

"Lost in follow‐up: 7/130 vs 5/130, wound dehiscence: 16/130 vs 10/130". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 7/130 vs 5/130, wound dehiscence: 16/130 vs 10/130". Extreme case: 23/130 vs 10/130, P=0.024; 16/130 vs 15/130, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Lost in follow‐up: 7/130 vs 5/130, P=0.769; wound dehiscence: 16/130 vs 10/130, P=0.301". "During the follow‐up period, 7 patients, who previously had undergone primary closure, were lost to follow‐up evaluation, of these 2 with normal findings were not motivated to continue their check‐ups. In the Limberg flap arm, 5 patients were lost at follow‐up evaluation, of these, 1 with normal findings was not motivated to continue their check‐ups". No difference between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. An independent observer, who was not from the surgical team and who was unaware of the treatment assignments, recorded all data, which included postoperative events and follow‐up findings.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.7.2 Modified Limberg flap

Galal 2012

Low risk of bias

Randomization was achieved through a computer‐generated schedule and the results were sealed into envelopes. The envelopes were drawn and opened by a nurse in the operating room. There was no significant difference between both groups regarding age, sex, preoperative symptoms, and period of follow‐up.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Chi square test, and Fisher’s exact test was used for qualitative data".

Some concerns

Based on judgements of previous domains.

Subgroup 1.7.3 Karydakis flap

Sakr 2006

Low risk of bias

Written consent was obtained from all patients who were categorized, using the closed envelope method, into 2 groups. “All baseline differences were not statistically significant”.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.7.4 Rotational flap

Enshaei 2014

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. According to the chi‐square test, there is no significant difference between sexes in two surgical procedures (P=0.1). The mean age of the patients was 24.17± 6.05 minimum 15 and maximum 40.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

"Depending on the clinical status, patients discharged and the first and fourth weeks and 6 months postoperatively were visited and questionnaires were completed". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Then, depending on the clinical status, patients discharged and the first and fourth weeks and 6 months postoperatively were visited and questionnaires were completed."

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 1.8 Time to return to work.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 1.8.1 Limberg flap

Akca 2005

Low risk of bias

Randomisation method: A computer‐generated table of random numbers. The randomisation was carried out by an independent computer consultant. No imbalances are apparent in the baseline characteristics among the randomized groups.

Low risk of bias

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

“Duration of incapacity for work was defined as the time from the date of surgery to the date on which the patient returned to normal activities including employment and leisure activities.” In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Dass 2012

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: mean 28.4±8.6 years (range 16–48 years). Gender: Intervention group: 38 male, 2 female. Control group: 36 male, 4 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Maghsudi 2020

Low risk of bias

A block randomization method (including qua druple blocks) was used for this purpose. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 24.0 ± 4.6 (range 14–36); control group: 24.1 ± 4.5 (range 14–34); Gender: Intervention group: 29(58%) male, 21(42%) female; control group: 33(66%) male, 17(34%) female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Muzi 2010

Low risk of bias

Eligible patients were assigned randomly to 1 of the 2 surgical groups according to a computer‐generated list. Allocations were sealed in opaque numbered envelopes. The code was broken on the day of surgery before the induction of anesthetics. There was no significant difference between the groups regarding sex, age, and presence of acute infection.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

"Lost in follow‐up: 7/130 vs 5/130". For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Some concerns

The measurement method may be suitable. “An independent observer, who was not from the surgical team and who was unaware of the treatment assignments, recorded all data, which included postoperative events and follow‐up findings”. However, outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 1.8.2 Modified Limberg flap

Galal 2012

Low risk of bias

Randomization was achieved through a computer‐generated schedule and the results were sealed into envelopes. The envelopes were drawn and opened by a nurse in the operating room. There was no significant difference between both groups regarding age, sex, preoperative symptoms, and period of follow‐up.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

Duration of incapacity for work was defined as the time from the date of surgery to the date on which the patient returned to normal activities including employment and leisure activities. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Duration of incapacity for work was defined as the time from the date of surgery to the date on which the patient returned to normal activities including employment and leisure activities".

Some concerns

Based on judgements of previous domains.

Shabbir 2014

Low risk of bias

A total of 60 patients were randomly divided into two equal groups using the lottery method to prevent selection bias. Each group had 30(50%) patients. In group 1 there were 27(90%) males and 3(10%) females, while group 2 had 28(93.3%) males and 2(6.7%) females.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 2.1 Proportion of wounds healed at 3 months.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Hemmingsson 2022

Low risk of bias

The randomization sequence was computer‐generated using a random number technique and concealed. After informed consent was obtained, allocation to a study arm was provided online from a secure web‐based registration and randomization application. Permuted blocks were used, and randomization stratified by centre. The permuted block sizes were not disclosed to the participating clinics. The entire process of registration, randomization, and allocation was completed on the study homepage, with coding performed by a system developer independent of the remainder of the study recruitment, conduct, and analysis. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD): Intervention group: 26.5 ± 1.5 (range 23‐30); control group:27.7 ± 1.3(range 23‐29). Gender: Intervention group: 52 male, 8 female; control group: 51 male, 5 female. BMI (mean ± SD ): Intervention group: 27.1 ± 0.6 (range 26‐26); control group: 26.0 ± 0.4 (range 25‐27).

Low risk of bias

"Nurses responsible for postoperative wound care and registration were not involved in the surgery" The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

“Lost in follow‐up: 2/56 vs 0/60”.

Low risk of bias

Wound healing was evaluated at the outpatient clinic at 1‐ to 2‐week intervals until complete wound healing was documented. Complete wound healing was defined as a dry scar requiring no further care or dressing. Nurses responsible for postoperative wound care and registration were not involved in the surgery. Observer‐reported outcomes involving some judgement. A nonsurgical team member recorded the patient data.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT00412659 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The patients came for follow‐up visits on PODs 2, 10, and 30 and then once every 6 months. At the end of the study, all patients were contacted again to obtain recurrence information". "The v2 test and Fisher’s exact test were used for categoric data comparisons".

Low risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 2.3 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.3.1 Limberg flap

Arnous 2019

Low risk of bias

"Randomisation was achieved through computer software and the results were sealed into envelopes." No imbalances are apparent in the baseline characteristics among the randomized groups.

High risk of bias

“The trial was single blinded as the patients gave consent to participate in the study, yet they were not aware of the type of procedure they will undergo. The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation.” The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 1 eligible trial participants in each group post‐randomization. Recurrence: 6/30 vs 0/30. Lost in follow‐up: 1/31 vs 1/31. The number of participants (n=2) who were analysed excluded from the analysis may have a substantial impact on the result (Recurrence: 7/31 vs 0/31, P=0.011; Recurrence: 6/31 vs 1/31, P=0.104).

High risk of bias

“Recurrence: 6/30 vs 0/30. Lost in follow‐up: 1/31 vs 1/31.” The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 6/30 vs 0/30. Lost in follow‐up: 1/31 vs 1/31.” Extreme case: 7/31 vs 0/31, P=0.011; 6/31 vs 1/31, P=0.104. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Recurrence: 6/30 vs 0/30, P=0.024; Lost in follow‐up: 1/31 vs 1/31.". The effects of the experimental and comparator interventions on the outcome are different. The proportion of participants with missing data is unlikely to differ between the intervention groups.

Low risk of bias

“The primary outcome was the recurrence of SPND at 12 months postoperatively. Recurrence was defined as reappearance of pilonidal sinus at the site of surgery as diagnosed by physical examination by the operating surgeon and a surgical resident.” "The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation."

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at www.researchregistry.com (researchregistry3601). The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Okuş 2012

Low risk of bias

Patients were randomized into two groups by the closed envelope method. The patients who expressed preference for one technique were not included in the study. No difference between groups for age and sex.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 7 eligible trial participants post‐randomization. Recurrence: 2/49 vs 2/44. Lost in follow‐up: 1/50 vs 6/50. The number of participants (n=7) who were analysed excluded from the analysis was not sufficient that there could have been a substantial impact on the result (Recurrence: 2/50 vs 8/50, P=0.091; Recurrence: 3/50 vs 2/50, P=0.999).

Low risk of bias

"Lost in follow‐up: 1/50 vs 6/50, Recurrence rate: 2/49 vs 2/44". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 1/50 vs 6/50, Recurrence rate: 2/49 vs 2/44". Extreme case: 3/50 vs 2/50, P=0.999; 2/50 vs 8/50, P=0.092. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "During the first 6 months of follow‐up there was no recurrence".

Some concerns

Based on judgements of previous domains.

Rashidian 2014

Some concerns

There is no information available regarding the allocation concealment in the study. Lack of information on baseline differences between groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Some concerns

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Tavassoli 2011

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. Lack of information on baseline differences between groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Long term follow‐up after 6 months was performed".

Some concerns

Based on judgements of previous domains.

Subgroup 2.3.2 Karydakis flap

Hemmingsson 2022

Low risk of bias

The randomization sequence was computer‐generated using a random number technique and concealed. After informed consent was obtained, allocation to a study arm was provided online from a secure web‐based registration and randomization application. Permuted blocks were used, and randomization stratified by centre. The permuted block sizes were not disclosed to the participating clinics. The entire process of registration, randomization, and allocation was completed on the study homepage, with coding performed by a system developer independent of the remainder of the study recruitment, conduct, and analysis. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD): Intervention group: 26.5 ± 1.5 (range 23‐30); control group:27.7 ± 1.3(range 23‐29). Gender: Intervention group: 52 male, 8 female; control group: 51 male, 5 female. BMI (mean ± SD ): Intervention group: 27.1 ± 0.6 (range 26‐26); control group: 26.0 ± 0.4 (range 25‐27).

Low risk of bias

"Nurses responsible for postoperative wound care and registration were not involved in the surgery" The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

"Lost in follow‐up: 5/56 vs 3/60, Recurrence rate: 5/56 vs 5/60". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 5/56 vs 3/60, Recurrence rate: 5/56 vs 5/60". Extreme case: 10/56 vs 5/56, P=0.267; 5/56 vs 8/56, P=0.557. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Low risk of bias

The time elapsed from surgery to the day of suture removal was defined as the shortest possible wound healing time if no wound complications occurred within 30 days. "Nurses responsible for postoperative wound care and registration were not involved in the surgery". Observer‐reported outcomes involving some judgement.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT00412659 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The patients came for follow‐up visits on PODs 2, 10, and 30 and then once every 6 months. At the end of the study, all patients were contacted again to obtain recurrence information". "The v2 test and Fisher’s exact test were used for categoric data comparisons".

Low risk of bias

Based on judgements of previous domains.

Subgroup 2.3.3 Limberg flap & Karydakis flap

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"The patients were informed that they would receive annual phone calls to follow up on their recurrences and further complaints of PD. After the first year after surgery, the patients were called by phone annually." "The patients were called by phone annually". Outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 2.7 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.7.1 Limberg flap

Arnous 2019

Low risk of bias

"Randomisation was achieved through computer software and the results were sealed into envelopes." No imbalances are apparent in the baseline characteristics among the randomized groups.

High risk of bias

“The trial was single blinded as the patients gave consent to participate in the study, yet they were not aware of the type of procedure they will undergo. The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation.” The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 1 eligible trial participants in each group post‐randomization. Wound infection: 6/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. The number of participants (n=2) who were analysed excluded from the analysis may have a substantial impact on the result (Wound infection: 7/31 vs 1/31, P=0.053; Recurrence: 6/31 vs 2/31, P=0.255).

Low risk of bias

“Wound infection: 6/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 6/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. ” Extreme case: 7/31 vs 1/31, P=0.053; 6/31 vs 2/31, P=0.255. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect.

Low risk of bias

"idefined as leakage of purulent secretion through the surgical wound" "The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation". Observer‐reported outcomes involving some judgement.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at www.researchregistry.com (researchregistry3601). The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Okuş 2012

Low risk of bias

Patients were randomized into two groups by the closed envelope method. The patients who expressed preference for one technique were not included in the study. No difference between groups for age and sex.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 7 eligible trial participants post‐randomization. Wound infection: 1/49 vs 1/44. Lost in follow‐up: 1/50 vs 6/50. The number of participants (n=5) who were analysed excluded from the analysis was not sufficient that there could have been a substantial impact on the result (Wound infection: 1/50 vs 7/50, P=0.059; Wound infection: 2/50 vs 1/50, P=0.999).

Low risk of bias

"Lost in follow‐up: 1/50 vs 6/50, wound infection: 1/49 vs 1/44". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 1/50 vs 6/50, wound infection: 1/49 vs 1/44". Extreme case: 2/50 vs 1/50, P=0.999; 1/50 vs 7/50, P=0.059. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. The observed number of events is not much greater than the number of participants with missing outcome data. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received. There was no pre‐registered protocol or statistical plan for this study.

Rashidian 2014

Some concerns

There is no information available regarding the allocation concealment in the study. Lack of information on baseline differences between groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"While careful physical examination by the surgeon evaluated progress of the wound healing and complications such as wound infection, dehiscence, flap necrosis, seroma, haematoma and bleeding". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

There is no information available regarding the allocation concealment in the study. Lack of information on baseline differences between groups. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses. The outcome is measured in all participants In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received. There was no pre‐registered protocol or statistical plan for this study.

Tavassoli 2011

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. Lack of information on baseline differences between groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. Lack of information on baseline differences between groups. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis. The outcome is measured in all participants The measurement method may be suitable. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received. There was no pre‐registered protocol or statistical plan for this study.

Subgroup 2.7.2 Karydakis flap

Hemmingsson 2022

Low risk of bias

The randomization sequence was computer‐generated using a random number technique and concealed. After informed consent was obtained, allocation to a study arm was provided online from a secure web‐based registration and randomization application. Permuted blocks were used, and randomization stratified by centre. The permuted block sizes were not disclosed to the participating clinics. The entire process of registration, randomization, and allocation was completed on the study homepage, with coding performed by a system developer independent of the remainder of the study recruitment, conduct, and analysis. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD): Intervention group: 26.5 ± 1.5 (range 23‐30); control group:27.7 ± 1.3(range 23‐29). Gender: Intervention group: 52 male, 8 female; control group: 51 male, 5 female. BMI (mean ± SD ): Intervention group: 27.1 ± 0.6 (range 26‐26); control group: 26.0 ± 0.4 (range 25‐27).

Low risk of bias

"Nurses responsible for postoperative wound care and registration were not involved in the surgery" The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Some concerns

"Lost in follow‐up: 2/56 vs 0/60, wound infection: 4/56 vs 0/60". The observed number of events is not much greater than the number of participants with missing outcome data. "Lost in follow‐up: 2/56 vs 0/60, wound infection: 4/56 vs 0/60". Extreme case: 6/56 vs 0/60, P=0.028; 4/56 vs 0/56, P=0.118. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Lost in follow‐up: 2/56 vs 0/60, P=0.231, wound infection: 4/56 vs 0/60, P=0.051". No difference between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

"Southampton Wound Assessment Scale 13". "Nurses responsible for postoperative wound care and registration were not involved in the surgery". Observer‐reported outcomes involving some judgement.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT00412659 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "Wound Assessment Scale was treated as an ordinal variable". "Differences in proportions between groups were tested with Fisher’s two‐sided exact test".

Some concerns

Based on judgements of previous domains.

Subgroup 2.7.3 Limberg flap & Karydakis flap

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants

Some concerns

Infection was defined as erythema and swelling on the wound side with purulent drainage. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis. The outcome is measured in all participants In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received. There was no pre‐registered protocol or statistical plan for this study.

Risk of bias for analysis 2.10 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 2.10.1 Limberg flap

Arnous 2019

Low risk of bias

"Randomisation was achieved through computer software and the results were sealed into envelopes." No imbalances are apparent in the baseline characteristics among the randomized groups.

Some concerns

“The trial was single blinded as the patients gave consent to participate in the study, yet they were not aware of the type of procedure they will undergo. The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation.” The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 2 eligible trial participants in each group post‐randomization. Wound dehiscence: 2/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. The number of participants (n=2) who were analysed excluded from the analysis was not sufficient that there could have been a substantial impact on the result (Wound dehiscence: 2/31 vs 2/31, P=1.0; Recurrence: 3/31 vs 1/31, P=0.612).

Low risk of bias

“Wound dehiscence: 2/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. ” The number of participants (n=2) with missing outcome data could have no important difference to the estimated effect of intervention. “Wound dehiscence: 2/30 vs 1/30. Lost in follow‐up: 1/31 vs 1/31. ” Extreme case: 3/31 vs 1/31, P=0.612; 2/31 vs 2/31, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect.

Low risk of bias

The measurement method may be suitable. "The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation." Observer‐reported outcomes not involving judgement.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at www.researchregistry.com (researchregistry3601). The results were unlikely to have been selected from multiple eligible outcome measurements.

Some concerns

Based on judgements of previous domains.

Rashidian 2014

Some concerns

There is no information available regarding the allocation concealment in the study. Lack of information on baseline differences between groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

"While careful physical examination by the surgeon evaluated progress of the wound healing and complications such as wound infection, dehiscence, flap necrosis, seroma, haematoma and bleeding". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 2.10.2 Limberg flap & Karydakis flap

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.1 Time to wound healing.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Alvandipour 2019

Some concerns

"The patients were assigned to 2 groups, i.e., the Limberg flap (LF; 27 patients) and the Karydakis flap (KF; 37 patients) groups, by using a computer‐based table of randomization to draw sealed envelopes randomly." Age: Intervention group: mean 25.89 ± 9.48 years; control group: mean 34.19 ± 10.47 years; (P = 0.002).

Some concerns

"In this single‐blind study, the participants were familiar with two common types of IPD surgery, but due to skin wound dressing on the back, they were not aware of their treatment. In addition, the observer, who completed the questionnaire, was not familiar with the groups. Similarly, the data analyzer did not know the study groups. However, the main researcher (surgeon) had information about the groups, and both surgical procedures were performed by the same surgeon. The safety and data monitoring committee was also informed about the study groups." The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. “One patient was lost during follow‐up due to family problems”. Analyses excluding 1 eligible trial participant post‐randomization. Lost in follow‐up: 0/37 vs 1/28. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Low risk of bias

“Lost in follow‐up: 0/37 vs 1/28”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient. The proportion of missing outcome data is sufficiently low that any bias is too small to be of importance.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study.

Some concerns

Based on judgements of previous domains.

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

Some concerns

“Lost in follow‐up: 19/110 vs 14/110”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). No differences between intervention groups in the proportions of missing outcome data (19/110 vs 14/110, P=0.451). No difference between the effects of the experimental and comparator interventions on the outcome (30.9 ± 10.7, N=91 vs 29.0 ± 9.9, N=96; P=0.209).

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study.

High risk of bias

Based on judgements of previous domains.

Bali 2015

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02287935).

Some concerns

Based on judgements of previous domains.

El Hadidi 2019

Low risk of bias

The randomization results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 23.1 ± 4.8 (median, range: 22, 18–40); control group: 22.8 ± 3.5 (median, range: 22, 19–35). Gender: Intervention group: 57 male, 3 female; control group: 58 male, 2 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

“Lost in follow‐up: 4/60 vs 5/60”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The t‐test was used to compare the qualitative data of the two groups. Paired sample t‐test was used to compare one group at different times.

Some concerns

Based on judgements of previous domains.

Elhiny 2021

Low risk of bias

An independent clinician was responsible for generating a random allocation sequence that kept hidden from all the trial participants. Based on type of operation, the participants were assigned equally into two groups using a computer‐generated block randomization technique. The printed letters were folded several times and placed within an aluminum foil. The letter placed inside an opaque sealed and stapled envelope, including patient’s identification code, name, time, and date. The type of treatment was masked for the patient (singleblinded technique). No statistically significant differences were observed between the groups regarding age and sex.

Low risk of bias

"The type of treatment was masked for the patient (singleblinded technique). " The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The Student’s t test was used to compare two means.

Some concerns

Based on judgements of previous domains.

Kohla 2016

Some concerns

Thirty patients with pilonidal disease were included in this prospective study from April 2012 to December 2013 and were divided randomly into two groups. There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: range 18–25; control group: range 19–32 Gender: not mentioned. BMI (mean ± SD ): Intervention group: six patients were obese and four patients were overweight; control group: four patients were overweight, and five patients were obese.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. Fisher’s exact test: This is used to compare between two groups or more for one qualitative variable in a 2 × 2 contingency table.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.3 Proportion of wounds healed.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Kohla 2016

Some concerns

Thirty patients with pilonidal disease were included in this prospective study from April 2012 to December 2013 and were divided randomly into two groups. There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: range 18–25; control group: range 19–32 Gender: not mentioned. BMI (mean ± SD ): Intervention group: six patients were obese and four patients were overweight; control group: four patients were overweight, and five patients were obese.

Low risk of bias

In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.4 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Subgroup 3.4.1 Follow‐up ＜12 months

Alvandipour 2019

Some concerns

"The patients were assigned to 2 groups, i.e., the Limberg flap (LF; 27 patients) and the Karydakis flap (KF; 37 patients) groups, by using a computer‐based table of randomization to draw sealed envelopes randomly." Age: Intervention group: mean 25.89 ± 9.48 years; control group: mean 34.19 ± 10.47 years; (P = 0.002).

Some concerns

"In this single‐blind study, the participants were familiar with two common types of IPD surgery, but due to skin wound dressing on the back, they were not aware of their treatment. In addition, the observer, who completed the questionnaire, was not familiar with the groups. Similarly, the data analyzer did not know the study groups. However, the main researcher (surgeon) had information about the groups, and both surgical procedures were performed by the same surgeon. The safety and data monitoring committee was also informed about the study groups." The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. “One patient was lost during follow‐up due to family problems”. Analyses excluding 1 eligible trial participant post‐randomization. Recurrence: 0/27 vs 1/37. Lost in follow‐up: 0/37 vs 1/28. The number of participants (n=1) who were analysed excluded from the analysis, was not sufficient that there could have been a substantial impact on the result.

Low risk of bias

“Recurrence: 0/27 vs 1/37. Lost in follow‐up: 1/28 vs 0/37”. The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 0/27 vs 1/37. Lost in follow‐up: 1/28 vs 0/37”. Extreme case: 1/28 vs 1/37, P=0.999; 0/28 vs 1/37, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make no important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The recurrence rate was evaluated at 6 months postsurgery".

Some concerns

Based on judgements of previous domains.

Anwar 2021

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 1 eligible trial participant post‐randomization. Recurrence: 0/35 vs 1/34. Lost in follow‐up: 0/35 vs 1/35. The number of participants (n=1) who were analysed excluded from the analysis, was not sufficient that there could have been a substantial impact on the result.

Low risk of bias

“Recurrence: 0/35 vs 1/34. Lost in follow‐up: 0/35 vs 1/35.” The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 0/35 vs 1/34. Lost in follow‐up: 0/35 vs 1/35.” Extreme case: 0/35 vs 1/35, P=0.999; 0/35 vs 2/35, P=0.493. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make no important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "All patients were followed for six months".

Some concerns

Based on judgements of previous domains.

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

Some concerns

“Recurrence: 10/91 vs 6/96. Lost in follow‐up: 19/110 vs 14/110.” The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 10/91 vs 6/96. Lost in follow‐up: 19/110 vs 14/110.” Extreme case: 29/110 vs 6/110, P=0.001; 10/110 vs 20/110, P=0.076. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Recurrence: 10/91 vs 6/96, P=0.301. Lost in follow‐up: 19/110 vs 14/110, P=0.451.” No difference between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Some concerns

The measurement method may be suitable. "presence of PSD was defined as recurrence" No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Riaz 2019

Low risk of bias

Patients were randomly divided into two equal groups of n patients each by lottery method. The overall mean age of patients included in our study was 27.21±8.24 years. The mean age of patients in Group L was28.06±8.71 years while in Group K, the mean age of patients was 26.36±7.69 years (p=0.142). The study comprised predominantly of male patients with the male to female ratio being 1:0.23. The distribution of patients according to gender is depicted in Table 1. The difference between the two groups was not significant (p=0.281).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Subgroup 3.4.2 Follow‐up ≥12 months

Bali 2015

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02287935). The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

El Hadidi 2019

Low risk of bias

The randomization results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 23.1 ± 4.8 (median, range: 22, 18–40); control group: 22.8 ± 3.5 (median, range: 22, 19–35). Gender: Intervention group: 57 male, 3 female; control group: 58 male, 2 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

“Lost in follow‐up: 4/60 vs 5/60, recurrence: 0/60 vs 0/60". The observed number of events is not much greater than the number of participants with missing outcome data. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Elhiny 2021

Low risk of bias

An independent clinician was responsible for generating a random allocation sequence that kept hidden from all the trial participants. Based on type of operation, the participants were assigned equally into two groups using a computer‐generated block randomization technique. The printed letters were folded several times and placed within an aluminum foil. The letter placed inside an opaque sealed and stapled envelope, including patient’s identification code, name, time, and date. The type of treatment was masked for the patient (singleblinded technique). No statistically significant differences were observed between the groups regarding age and sex.

Low risk of bias

"The type of treatment was masked for the patient (singleblinded technique). " The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. "The Chi‐square test was used to compare between two or more proportions".

Some concerns

Based on judgements of previous domains.

Kohla 2016

Some concerns

Thirty patients with pilonidal disease were included in this prospective study from April 2012 to December 2013 and were divided randomly into two groups. There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: range 18–25; control group: range 19–32 Gender: not mentioned. BMI (mean ± SD ): Intervention group: six patients were obese and four patients were overweight; control group: four patients were overweight, and five patients were obese.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"The patients were informed that they would receive annual phone calls to follow up on their recurrences and further complaints of PD. After the first year after surgery, the patients were called by phone annually." "The patients were called by phone annually". Outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.6 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Alvandipour 2019

Some concerns

"The patients were assigned to 2 groups, i.e., the Limberg flap (LF; 27 patients) and the Karydakis flap (KF; 37 patients) groups, by using a computer‐based table of randomization to draw sealed envelopes randomly." Age: Intervention group: mean 25.89 ± 9.48 years; control group: mean 34.19 ± 10.47 years; (P = 0.002).

Some concerns

"In this single‐blind study, the participants were familiar with two common types of IPD surgery, but due to skin wound dressing on the back, they were not aware of their treatment. In addition, the observer, who completed the questionnaire, was not familiar with the groups. Similarly, the data analyzer did not know the study groups. However, the main researcher (surgeon) had information about the groups, and both surgical procedures were performed by the same surgeon. The safety and data monitoring committee was also informed about the study groups." The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. “One patient was lost during follow‐up due to family problems”. Analyses excluding 1 eligible trial participant post‐randomization. Wound infection: 1/27 vs 3/37. Lost in follow‐up: 0/37 vs 1/28. The number of participants (n=1) who were analysed excluded from the analysis, was not sufficient that there could have been a substantial impact on the result.

Low risk of bias

“Wound infection: 1/27 vs 3/37. Lost in follow‐up: 1/28 vs 0/37. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 1/27 vs 3/37. Lost in follow‐up: 1/28 vs 0/37. ” Extreme case: 2/28 vs 3/37, P=0.999; 1/28 vs 3/37, P=0.632. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make no important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Anwar 2021

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 1 eligible trial participant post‐randomization. Wound infection: 2/35 vs 2/34. Lost in follow‐up: 0/35 vs 1/35. The number of participants (n=1) who were analysed excluded from the analysis, was not sufficient that there could have been a substantial impact on the result.

Low risk of bias

“Wound infection: 2/35 vs 2/34. Lost in follow‐up: 0/35 vs 1/35. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 2/35 vs 2/34. Lost in follow‐up: 0/35 vs 1/35. ” Extreme case: 2/35 vs 2/35, P=0.999; 2/35 vs 3/35, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make no important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

Some concerns

“Wound infection: 6/91 vs 2/96. Lost in follow‐up: 19/110 vs 14/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 6/91 vs 2/96. Lost in follow‐up: 19/110 vs 14/110. ” Extreme case: 25/110 vs 2/110, P=0.001; 6/110 vs 16/110, P=0.041. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound infection: 6/91 vs 2/96, P=0.161. Lost in follow‐up: 19/110 vs 14/110, P=0.451. ” No difference between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. "The distinction between seroma and infection was made based on the clinical findings and culture results." No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Culture result is an automated test. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Bali 2015

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02287935). There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

El Hadidi 2019

Low risk of bias

The randomization results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 23.1 ± 4.8 (median, range: 22, 18–40); control group: 22.8 ± 3.5 (median, range: 22, 19–35). Gender: Intervention group: 57 male, 3 female; control group: 58 male, 2 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

“Lost in follow‐up: 4/60 vs 5/60, Wound infection: 2/60 vs 1/60". The observed number of events is not much greater than the number of participants with missing outcome data. “Lost in follow‐up: 4/60 vs 5/60, Wound infection: 2/60 vs 1/60". Extreme case: 6/60 vs 5/60, P=0.999; 4/60 vs 6/60, P=0.743. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect. "Nine patients could not be followed up because they moved aboard". All missing outcome data occurred for documented reasons that are unrelated to the outcome.

Some concerns

"Postoperative infection was defined as the development of cellulitis and/or purulent discharge from the wound edges or drains. " No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "χ2‐test was used for qualitative data".

Some concerns

Based on judgements of previous domains.

Elhiny 2021

Low risk of bias

An independent clinician was responsible for generating a random allocation sequence that kept hidden from all the trial participants. Based on type of operation, the participants were assigned equally into two groups using a computer‐generated block randomization technique. The printed letters were folded several times and placed within an aluminum foil. The letter placed inside an opaque sealed and stapled envelope, including patient’s identification code, name, time, and date. The type of treatment was masked for the patient (singleblinded technique). No statistically significant differences were observed between the groups regarding age and sex.

Low risk of bias

"The type of treatment was masked for the patient (singleblinded technique). " The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The Chi‐square test was used to compare between two or more proportions".

Some concerns

Based on judgements of previous domains.

Riaz 2019

Low risk of bias

Patients were randomly divided into two equal groups of n patients each by lottery method. The overall mean age of patients included in our study was 27.21±8.24 years. The mean age of patients in Group L was28.06±8.71 years while in Group K, the mean age of patients was 26.36±7.69 years (p=0.142). The study comprised predominantly of male patients with the male to female ratio being 1:0.23. The distribution of patients according to gender is depicted in Table 1. The difference between the two groups was not significant (p=0.281).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

Infection was defined as erythema and swelling on the wound side with purulent drainage. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.8 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Alvandipour 2019

Some concerns

"The patients were assigned to 2 groups, i.e., the Limberg flap (LF; 27 patients) and the Karydakis flap (KF; 37 patients) groups, by using a computer‐based table of randomization to draw sealed envelopes randomly." Age: Intervention group: mean 25.89 ± 9.48 years; control group: mean 34.19 ± 10.47 years; (P = 0.002).

Some concerns

"In this single‐blind study, the participants were familiar with two common types of IPD surgery, but due to skin wound dressing on the back, they were not aware of their treatment. In addition, the observer, who completed the questionnaire, was not familiar with the groups. Similarly, the data analyzer did not know the study groups. However, the main researcher (surgeon) had information about the groups, and both surgical procedures were performed by the same surgeon. The safety and data monitoring committee was also informed about the study groups." The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. “One patient was lost during follow‐up due to family problems”. Analyses excluding 1 eligible trial participant post‐randomization. Wound dehiscence: 0/27 vs 1/37. Lost in follow‐up: 0/37 vs 1/28. The number of participants (n=1) who were analysed excluded from the analysis, was not sufficient that there could have been a substantial impact on the result.

Low risk of bias

“Wound dehiscence: 0/27 vs 1/37. Lost in follow‐up: 1/28 vs 0/37. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound dehiscence: 0/27 vs 1/37. Lost in follow‐up: 1/28 vs 0/37. ” Extreme case: 1/28 vs 1/37, P=0.999; 0/28 vs 1/37, P=0.999. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make no important difference to the estimated intervention effect.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Wound dehiscence: 4/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound dehiscence: 4/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” Extreme case: 6/110 vs 2/110, P=0.280; 4/110 vs 16/110, P=0.008. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound dehiscence: 4/108 vs 2/96, P=0.686. Lost in follow‐up: 2/110 vs 14/110, P=0.003.. ” Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

High risk of bias

Based on judgements of previous domains.

Bali 2015

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02287935). The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

El Hadidi 2019

Low risk of bias

The randomization results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 23.1 ± 4.8 (median, range: 22, 18–40); control group: 22.8 ± 3.5 (median, range: 22, 19–35). Gender: Intervention group: 57 male, 3 female; control group: 58 male, 2 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

“Lost in follow‐up: 4/60 vs 5/60, Wound dehiscence: 2/60 vs 1/60". The observed number of events is not much greater than the number of participants with missing outcome data. “Lost in follow‐up: 4/60 vs 5/60, Wound dehiscence: 2/60 vs 1/60". Extreme case: 6/60 vs 5/60, P=0.999; 4/60 vs 6/60, P=0.743. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect. "Nine patients could not be followed up because they moved aboard". All missing outcome data occurred for documented reasons that are unrelated to the outcome.

Low risk of bias

"The wounds were inspected on days 3 and 6 after the operation." No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "χ2‐test was used for qualitative data".

Some concerns

Based on judgements of previous domains.

Kohla 2016

Some concerns

Thirty patients with pilonidal disease were included in this prospective study from April 2012 to December 2013 and were divided randomly into two groups. There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: range 18–25; control group: range 19–32 Gender: not mentioned. BMI (mean ± SD ): Intervention group: six patients were obese and four patients were overweight; control group: four patients were overweight, and five patients were obese.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Riaz 2019

Low risk of bias

Patients were randomly divided into two equal groups of n patients each by lottery method. The overall mean age of patients included in our study was 27.21±8.24 years. The mean age of patients in Group L was28.06±8.71 years while in Group K, the mean age of patients was 26.36±7.69 years (p=0.142). The study comprised predominantly of male patients with the male to female ratio being 1:0.23. The distribution of patients according to gender is depicted in Table 1. The difference between the two groups was not significant (p=0.281).

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Sevinc 2016

Low risk of bias

The patients were randomized into 3 groups by the closed‐envelope method in the operating room. The groups were similar in terms of age, sex, American Society of Anesthesiologists score and body mass index.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 3.10 Time to return to work.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Alvandipour 2019

Some concerns

"The patients were assigned to 2 groups, i.e., the Limberg flap (LF; 27 patients) and the Karydakis flap (KF; 37 patients) groups, by using a computer‐based table of randomization to draw sealed envelopes randomly." Age: Intervention group: mean 25.89 ± 9.48 years; control group: mean 34.19 ± 10.47 years; (P = 0.002).

Some concerns

"In this single‐blind study, the participants were familiar with two common types of IPD surgery, but due to skin wound dressing on the back, they were not aware of their treatment. In addition, the observer, who completed the questionnaire, was not familiar with the groups. Similarly, the data analyzer did not know the study groups. However, the main researcher (surgeon) had information about the groups, and both surgical procedures were performed by the same surgeon. The safety and data monitoring committee was also informed about the study groups." The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. “One patient was lost during follow‐up due to family problems”. Analyses excluding 1 eligible trial participant post‐randomization. Lost in follow‐up: 0/37 vs 1/28. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Low risk of bias

“Lost in follow‐up: 0/37 vs 1/28”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient. The proportion of missing outcome data is sufficiently low that any bias is too small to be of importance.

Some concerns

The measurement method may be suitable. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Anwar 2021

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 1 eligible trial participant post‐randomization. “Lost in follow‐up: 0/35 vs 1/35”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Low risk of bias

“Lost in follow‐up: 0/35 vs 1/35”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed

Some concerns

Based on judgements of previous domains.

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Lost in follow‐up: 19/110 vs 14/110”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). No differences between intervention groups in the proportions of missing outcome data (19/110 vs 14/110, P=0.451). Differences between the effects of the experimental and comparator interventions on the outcome (19.1 ± 3.4, N=91 vs 20.8 ± 6.5, N=96; P=0.027).

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Bali 2015

Some concerns

There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02287935). There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

El Hadidi 2019

Low risk of bias

The randomization results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. No imbalances are apparent in the baseline characteristics among the randomized groups. Age (mean ± SD ): Intervention group: 23.1 ± 4.8 (median, range: 22, 18–40); control group: 22.8 ± 3.5 (median, range: 22, 19–35). Gender: Intervention group: 57 male, 3 female; control group: 58 male, 2 female.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

“Lost in follow‐up: 4/60 vs 5/60”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Some concerns

"To define the postoperative time‐off work, the patients were asked to start their daily activities when they were comfortable. " No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Kohla 2016

Some concerns

Thirty patients with pilonidal disease were included in this prospective study from April 2012 to December 2013 and were divided randomly into two groups. There is no information available regarding the randomization process and allocation concealment in the study. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: range 18–25; control group: range 19–32 Gender: not mentioned. BMI (mean ± SD ): Intervention group: six patients were obese and four patients were overweight; control group: four patients were overweight, and five patients were obese.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 4.1 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Guner 2013

Low risk of bias

A computer‐assisted random number table. A nonsurgical team member recorded the patient data. Both groups showed a homogeneous distribution in terms of age, sex, BMI, duration of complaints, CF class, and previous treatments.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The term ‘‘recurrence’’ was used when symptoms of the disease recurred some time after complete wound healing. "A nonsurgical team member recorded the patient data". Observer‐reported outcomes involving some judgement.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT01337869 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The patients came for follow‐up visits on PODs 2, 10, and 30 and then once every 6 months. At the end of the study, all patients were contacted again to obtain recurrence information". "The v2 test and Fisher’s exact test were used for categoric data comparisons".

Low risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 4.2 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Guner 2013

Low risk of bias

A computer‐assisted random number table. A nonsurgical team member recorded the patient data. Both groups showed a homogeneous distribution in terms of age, sex, BMI, duration of complaints, CF class, and previous treatments.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

"Fluid collections (seroma and hematoma), wound dehiscence (partial and total), and surgical‐site infection (superficial and deep) were identified as postoperative surgical area‐related complications". "A nonsurgical team member recorded the patient data". Observer‐reported outcomes involving some judgement.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT01337869 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The v2 test and Fisher’s exact test were used for categoric data comparisons".

Low risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 4.3 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Guner 2013

Low risk of bias

A computer‐assisted random number table. A nonsurgical team member recorded the patient data. Both groups showed a homogeneous distribution in terms of age, sex, BMI, duration of complaints, CF class, and previous treatments.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

Fluid collections (seroma and hematoma), wound dehiscence (partial and total), and surgical‐site infection (superficial and deep) were identified as postoperative surgical area‐related complications. "A nonsurgical team member recorded the patient data". Observer‐reported outcomes not involving judgement.

Low risk of bias

https://clinicaltrials.gov/ct2/show/NCT01337869 The results were unlikely to have been selected from multiple eligible outcome measurements or multiple eligible analyses of the data. "The v2 test and Fisher’s exact test were used for categoric data comparisons".

Low risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 5.1 Time to wound healing.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Lost in follow‐up: 2/110 vs 14/110”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). Differences between intervention groups in the proportions of missing outcome data (2/110 vs 14/110, P=0.003).

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study.

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 5.2 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Recurrence: 2/108 vs 6/96. Lost in follow‐up: 2/110 vs 14/110.” The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 2/108 vs 6/96. Lost in follow‐up: 2/110 vs 14/110.” Extreme case: 4/110 vs 6/110, P=0.748; 2/110 vs 20/110, P=0.001. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Recurrence: 2/108 vs 6/96, P=0.152. Lost in follow‐up: 2/110 vs 14/110, P=0.003.” Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Some concerns

The measurement method may be suitable. "presence of PSD was defined as recurrence" No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 5.3 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Wound infection: 5/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 5/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” Extreme case: 7/110 vs 2/110, P=0.171; 5/110 vs 16/110, P=0.023. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound infection: 5/108 vs 2/96, P=0.451. Lost in follow‐up: 2/110 vs 14/110, P=0.003.. ” Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. "The distinction between seroma and infection was made based on the clinical findings and culture results." No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Culture result is an automated test. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 5.4 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Wound dehiscence: 4/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound dehiscence: 4/108 vs 2/96. Lost in follow‐up: 2/110 vs 14/110. ” Extreme case: 6/110 vs 2/110, P=0.280; 4/110 vs 16/110, P=0.008. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound dehiscence: 4/108 vs 2/96, P=0.686. Lost in follow‐up: 2/110 vs 14/110, P=0.003.. ” Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 5.5 Time to return to work.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Lost in follow‐up: 2/110 vs 14/110. ”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). Differences between intervention groups in the proportions of missing outcome data (2/110 vs 14/110, P=0.003).

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 6.1 Time to wound healing.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Lost in follow‐up: 2/110 vs 19/110”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). Differences between intervention groups in the proportions of missing outcome data (2/110 vs 19/110, P=0.001).

Some concerns

The methodology utilized for measuring the outcome of interest was not reported, leaving room for uncertainty regarding the appropriateness of the measurement method. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study.

High risk of bias

Based on judgements of previous domains.

Can 2010

Some concerns

Block randomization was done to ensure an equal number of patients assigned to each of the two treatment arms. Block sizes were randomly set between 4 and 8, inclusive, and were based on a computer‐generated random allocation scheme. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: median 22 (19–40) years; Control group: median 22 (20–31) years. Gender: Intervention group: 76 male, 1 female; Control group: 67 male, 1 female.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. 5 patients from Karydakis group changed to MLF group. However, trialists did not report whether deviations arose because of the trial context. Analyses excluding 8 eligible trial participants post‐randomization. Analyses excluding eligible trial participants post‐randomization should be considered inappropriate. Lost in follow‐up: 3/77 vs 5/68. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Low risk of bias

“Lost in follow‐up: 3/77 vs 5/68”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient. The proportion of missing outcome data is sufficiently low that any bias is too small to be of importance.

Some concerns

"Questionnaire used for data recording via telephone interview: When did you feel you were completely healed after the operation?" Outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. "The questionnaire used in the study for data recording involved a telephone interview, where patients were asked to report on postoperative day 30 when they felt completely healed after the operation".

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 6.2 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Recurrence: 2/108 vs 10/91. Lost in follow‐up: 2/110 vs 19/110.” The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 2/108 vs 10/91. Lost in follow‐up: 2/110 vs 19/110.” Extreme case: 4/110 vs 10/110, P=0.165; 2/110 vs 20/110, P=0.001. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Recurrence: 2/108 vs 10/91, P=0.013. Lost in follow‐up: 2/110 vs 19/110, P=0.001.” Differences between intervention groups in the proportions of missing outcome data. Differences between the effects of the experimental and comparator interventions on the outcome.

Some concerns

The measurement method may be suitable. "presence of PSD was defined as recurrence" No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Can 2010

Some concerns

Block randomization was done to ensure an equal number of patients assigned to each of the two treatment arms. Block sizes were randomly set between 4 and 8, inclusive, and were based on a computer‐generated random allocation scheme. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: median 22 (19–40) years; Control group: median 22 (20–31) years. Gender: Intervention group: 76 male, 1 female; Control group: 67 male, 1 female.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. 5 patients from Karydakis group changed to MLF group. However, trialists did not report whether deviations arose because of the trial context. Intention‐to‐treat analyses.

Low risk of bias

“Recurrence: 4/74 vs 3/63. Lost in follow‐up: 3/77 vs 5/68”. The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 4/74 vs 3/63. Lost in follow‐up: 3/77 vs 5/68”. Extreme case: 7/77 vs 3/68, P=0.336; 4/77 vs 8/68, P=0.227. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Some concerns

"All similar complaints that occurred following an uneventful recovery period were considered 'recurrent disease' ". "Have you experienced any problems with your surgical wound until now? If yes, at what time? 4 Did you have to see a doctor about your surgical site (postoperative)? If yes, were you diagnosed with recurrent pilonidal disease?" No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement or Participant‐reported outcomes. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. "Fisher’s exact tests were performed to compare the 2 groups for early failure, recurrence rate, and whether or not a given patient would recommend the same operation to others".

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 6.3 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Wound infection: 5/108 vs 6/91. Lost in follow‐up: 2/110 vs 19/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound infection: 5/108 vs 6/91. Lost in follow‐up: 2/110 vs 19/110. ” Extreme case: 7/110 vs 6/110, P=0.999; 5/110 vs 25/110, P=0.001. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound infection: 5/108 vs 6/91, P=0.554. Lost in follow‐up: 2/110 vs 19/110, P=0.001. ” Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. "The distinction between seroma and infection was made based on the clinical findings and culture results." No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Culture result is an automated test. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Can 2010

Some concerns

Block randomization was done to ensure an equal number of patients assigned to each of the two treatment arms. Block sizes were randomly set between 4 and 8, inclusive, and were based on a computer‐generated random allocation scheme. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: median 22 (19–40) years; Control group: median 22 (20–31) years. Gender: Intervention group: 76 male, 1 female; Control group: 67 male, 1 female.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. 5 patients from Karydakis group changed to MLF group. However, trialists did not report whether deviations arose because of the trial context. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Some concerns

Recognized during outpatient visit on postoperative day 30. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. Recognized during outpatient visit on postoperative day 30. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 6.4 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Wound dehiscence: 4/108 vs 14/91. Lost in follow‐up: 2/110 vs 19/110. ” The observed number of events is not much greater than the number of participants with missing outcome data. “Wound dehiscence: 4/108 vs 14/91. Lost in follow‐up: 2/110 vs 19/110. ” Extreme case: 6/110 vs 14/110, P=0.098; 4/110 vs 33/110, P=0.001. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. “Wound dehiscence: 4/108 vs 14/91, P=0.006. Lost in follow‐up: 2/110 vs 19/110, P=0.001. ” Differences between intervention groups in the proportions of missing outcome data. Differences between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Can 2010

Some concerns

Block randomization was done to ensure an equal number of patients assigned to each of the two treatment arms. Block sizes were randomly set between 4 and 8, inclusive, and were based on a computer‐generated random allocation scheme. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: median 22 (19–40) years; Control group: median 22 (20–31) years. Gender: Intervention group: 76 male, 1 female; Control group: 67 male, 1 female.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. 5 patients from Karydakis group changed to MLF group. However, trialists did not report whether deviations arose because of the trial context. Intention‐to‐treat analyses.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

"Recognized during the in‐hospital period and postoperative day 30". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. Recognized during the in‐hospital period and postoperative day 30. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 6.5 Time to return to work.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arslan 2014

Low risk of bias

"Block randomization" "For randomization, 123 envelopes were prepared for each group. An envelope designating the surgery type was selected just prior to the operation by the surgical nurse for each patient. The randomization was performed in the outpatient clinic." No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Modified Limberg flap: mean 24.7 ± 5.1 years; Karydakis flap: mean 24.7 ± 5.1 years; Limberg flap: mean 26.5 ± 5.9 years. Gender: Modified Limberg flap: 95 male, 13 female; Karydakis flap: 77 male, 14 female; Limberg flap: 68 male, 31 female.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 35 eligible trial participants post‐randomization. Lost in follow‐up: 14/110, 2/110, and 19/110. There could be a potential impact (more than 10% of participants lost follow‐up).

High risk of bias

“Lost in follow‐up: 2/110 vs 19/110.”. There may be a potential for a substantial impact (over 10% of participants were lost to follow up). There may be a potential for a substantial impact (over 10% of participants were lost to follow up). Differences between intervention groups in the proportions of missing outcome data (2/110 vs 19/110, P=0.001).

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Can 2010

Some concerns

Block randomization was done to ensure an equal number of patients assigned to each of the two treatment arms. Block sizes were randomly set between 4 and 8, inclusive, and were based on a computer‐generated random allocation scheme. No imbalances are apparent in the baseline characteristics among the randomized groups. Age: Intervention group: median 22 (19–40) years; Control group: median 22 (20–31) years. Gender: Intervention group: 76 male, 1 female; Control group: 67 male, 1 female.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. 5 patients from Karydakis group changed to MLF group. However, trialists did not report whether deviations arose because of the trial context. Intention‐to‐treat analyses.

Low risk of bias

“Lost in follow‐up: 3/77 vs 5/68”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient. The proportion of missing outcome data is sufficiently low that any bias is too small to be of importance.

Some concerns

"When did you feel you were ready to return to work after the operation?" No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. "When did you feel you were ready to return to work after the operation?" Student’s t test was used to compare continuous variable means for duration of operation, time to feeling ready to return to work.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 7.1 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Kartal 2019

Low risk of bias

Allocation sequence was random and allocation sequence was concealed. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT03424057). The results were unlikely to have been selected from multiple eligible outcome measurements.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 7.2 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Kartal 2019

Low risk of bias

Allocation sequence was random and allocation sequence was concealed. No imbalances are apparent.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT03424057). The results were unlikely to have been selected from multiple eligible outcome measurements.

Low risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 8.1 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Bessa 2013

Low risk of bias

"Pseudorandom number generator with individual assignments concealed in sequentially numbered sealed envelopes that were opened in order when assignments were made." No significant differences were observed between the groups regarding age, sex distribution, BMI, duration of complaint, or history of recurrent disease.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 5 eligible trial participants post‐randomization. Recurrence: 1/60 vs 2/60. Lost in follow‐up: 3/63 vs 2/62. The number of participants (n=5) who were analysed excluded from the analysis was not sufficient that there could have been a substantial impact on the result (Recurrence: 4/63 vs 2/62, P=0.680; Recurrence: 1/63 vs 3/62, P=0.365).

Low risk of bias

"Recurrence: 1/60 vs 2/60. Lost in follow‐up: 3/63 vs 2/62". The observed number of events is not much greater than the number of participants with missing outcome data. “Recurrence: 2/108 vs 10/91. Lost in follow‐up: 2/110 vs 19/110.” Extreme case: 4/63 vs 2/62, P=0.680; 1/63 vs 3/62, P=0.365 Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could not make important difference to the estimated intervention effect.

Some concerns

"Recurrent disease was reported if observed at any time throughout the study period." No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. "χ2 and the Fisher exact test were used for categorical variables".

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 8.2 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Bessa 2013

Low risk of bias

"Pseudorandom number generator with individual assignments concealed in sequentially numbered sealed envelopes that were opened in order when assignments were made." No significant differences were observed between the groups regarding age, sex distribution, BMI, duration of complaint, or history of recurrent disease.

Some concerns

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 5 eligible trial participants post‐randomization. Wound infection: 2/60 vs 3/60. Lost in follow‐up: 3/63 vs 2/62. The number of participants (n=5) who were analysed excluded from the analysis was not sufficient that there could have been a substantial impact on the result (Wound infection: 2/63 vs 4/62, P=0.440; Wound infection: 5/63 vs 3/62, P=0.717).

Low risk of bias

"Wound infection: 2/60 vs 3/60. Lost in follow‐up: 3/63 vs 2/62". The observed number of events is not much greater than the number of participants with missing outcome data. "Wound infection: 2/60 vs 3/60. Lost in follow‐up: 3/63 vs 2/62". Extreme case: 2/63 vs 4/62, P=0.440; 5/63 vs 3/62, P=0.717. Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 8.3 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Bessa 2013

Low risk of bias

"Pseudorandom number generator with individual assignments concealed in sequentially numbered sealed envelopes that were opened in order when assignments were made." No significant differences were observed between the groups regarding age, sex distribution, BMI, duration of complaint, or history of recurrent disease.

High risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 5 eligible trial participants post‐randomization. Wound disruption: 0/60 vs 9/60. Lost in follow‐up: 3/63 vs 2/62.There could be a potential impact (Wound infection: 3/63 vs 9/62, P=0.076; Wound infection: 0/63 vs 11/62, P=0.0003).

High risk of bias

"Wound disruption: 0/60 vs 9/60. Lost in follow‐up: 3/63 vs 2/62". The observed number of events is not much greater than the number of participants with missing outcome data. "Wound disruption: 0/60 vs 9/60. Lost in follow‐up: 3/63 vs 2/62". Extreme case: 3/63 vs 9/62, P=0.076; Wound infection: 0/63 vs 11/62, P=0.0003 Sensitivity analyses (conducted by the review authors) confirm that plausible values of the missing outcome data could make important difference to the estimated intervention effect. "Wound disruption: 0/60 vs 9/60, P=0.003. Lost in follow‐up: 3/63 vs 2/62, P=0.999". Differences between intervention groups in the proportions of missing outcome data. No difference between the effects of the experimental and comparator interventions on the outcome.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

High risk of bias

Based on judgements of previous domains.

Risk of bias for analysis 9.1 Recurrence rate.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Saydam 2015

Some concerns

The patients were randomly divided into two groups in the computer. No statistically siginificant differences were found in age, BMI, operative time and hospital length of stay between the groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

"The patients were asked to come for follow‐up appointments on the 10 th day especially for SSI, and first, sixth and twelfth month postoperatively. Any clinical changes such as recurrence, wound dehiscence, surgical site infection or any other any complication discovered during the follow‐up period was compared and recorded". No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02116738). There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 9.2 Wound infection.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Saydam 2015

Some concerns

The patients were randomly divided into two groups in the computer. No statistically siginificant differences were found in age, BMI, operative time and hospital length of stay between the groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Some concerns

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes involving some judgement. The assessment of outcome is potentially influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02116738). "The patients were asked to come for follow‐up appointments on the 10 th day especially for SSI". There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

Risk of bias for analysis 9.3 Wound dehiscence.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Saydam 2015

Some concerns

The patients were randomly divided into two groups in the computer. No statistically siginificant differences were found in age, BMI, operative time and hospital length of stay between the groups.

Low risk of bias

No information provided. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Intention‐to‐treat analysis.

Low risk of bias

The outcome is measured in all participants.

Low risk of bias

The measurement method may be suitable. No information provided. In the field of surgical research, it is possible for assessors to be aware of the assigned intervention of participants. Observer‐reported outcomes not involving judgement. The assessment of outcome is usually not likely to be influenced by knowledge of intervention received.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at clinicaltrials.gov (NCT02116738). There is only one possible way in which the outcome domain can be measured. There is only one possible way in which the outcome measurement can be analysed.

Some concerns

Based on judgements of previous domains.

The risk of bias brief characteristics of the included studies are as follows.

• Randomisation process. Most studies provided detailed descriptions of the randomisation process and allocation concealment, with no apparent baseline differences. Alvandipour 2019 demonstrated an apparent imbalance, particularly in baseline age, following the randomisation of participants. In some studies, there was no information available regarding the randomisation process and allocation concealment, with no apparent baseline differences (Anwar 2021;Bali 2015; Can 2010; Daglar 2004; Dass 2012; Enshaei 2014; Kohla 2016; Rashidian 2014; Saydam 2015; Tavassoli 2011).

• Deviations from intended interventions. In the field of surgical research, it is possible for participants, caregivers, and intervention providers to be aware of the assigned intervention of participants. However, the surgical procedure may not deviate significantly from the predetermined or intended strategy. Most studies utilised intention‐to‐treat analyses, but some studies excluded post‐randomisation patients from the analysis (Alvandipour 2019; Anwar 2021; Arnous 2019; Arslan 2014; Bessa 2013; Can 2010; Okuş 2012).

• Missing outcome data. Most studies included all or nearly all randomised participants.

• Measurement of the outcome. The majority of studies did not provide specific information about the methodology utilised for measuring the outcome of interest and blinding the outcome assessors. Based on our practical experience, outcomes such as time to wound healing, proportion of wounds healed, recurrence rate, wound infection, and time to return to work rely on participant or observer reports and involve some level of judgement. Unless study authors explicitly described methods to blind outcome assessors or other steps to make outcome assessments more objective, we consider the measurement of these outcomes to be of ‘some concern’.

• Selective reporting. Only two of the 33 included studies described a prospective clinical trial registration or statistical plan (Guner 2013; Hemmingsson 2022).

Effects of interventions

See: Table 1; Table 2; Table 3

1. Off‐midline closure versus conventional midline closure

Fourteen studies with 1614 analysed participants compared off‐midline closure to conventional midline closure (Abu 1999; Akca 2005; Bukhari 2020; Daglar 2004; Dass 2012; Enshaei 2014; Enshaei 2020; Ertan 2005; Galal 2012; Khan 2013; Maghsudi 2020; Muzi 2010; Sakr 2006; Shabbir 2014). See: Table 1.

1.1. Time to wound healing

Based on three studies with 300 participants (Enshaei 2014; Galal 2012; Maghsudi 2020), off‐midline closure probably reduces the time to wound healing compared to the conventional midline closure procedure (MD ‐5.23 days, 95% CI ‐7.55 to ‐2.92 days; I² = 92%; 3 studies, 300 participants; moderate‐certainty evidence; Analysis 1.1).

1.1. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 1: Time to wound healing

One other study offered additional insights: Ertan 2005 found that wound healing time for off‐midline closure with the Limberg flap was quicker at 10.3 days, compared to 15.7 days for the midline closure group.

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 0.02, degrees of freedom (df) = 1 (P = 0.88), I² = 0%; Analysis 1.1).

1.2. Proportion of wounds healed

Data were not available for the 1‐month, 6‐month, and 12‐month time points.

Based on two studies with 207 participants with nearly 10 days of follow‐up (Abu 1999; Sakr 2006), off‐midline closure may result in little to no difference in wound healing rates when compared to the conventional midline closure procedure (RR 1.13, 95% CI 0.92 to 1.39; I² = 66%; 2 studies, 207 participants; very low‐certainty evidence; Analysis 1.2). Additionally, Sakr 2006 reported that all patients experienced wound healing within 21 days.

1.2. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 2: Proportion of wounds healed

1.3. Recurrence rate

Based on 13 studies with 1492 participants with three months to 48 months of follow‐up (Abu 1999; Akca 2005; Daglar 2004; Dass 2012; Enshaei 2014; Enshaei 2020; Ertan 2005; Galal 2012; Khan 2013; Maghsudi 2020; Muzi 2010; Sakr 2006; Shabbir 2014), off‐midline closure probably results in lower rates of recurrence compared to the conventional midline closure procedure (RR 0.22, 95% CI 0.11 to 0.45; I² = 0%; 13 studies, 1492 participants; moderate‐certainty evidence; Analysis 1.3). We could not identify asymmetry in the funnel plot suggestive of publication bias (see Figure 2). Egger's test was not statistically significant (P = 0.5932).

1.3. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 3: Recurrence rate

2.

Funnel plot of comparison: 1 Off‐midline closure versus midline closure (Conventional), outcome: 1.3 Recurrence rate.

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 0.20, df = 3 (P = 0.98), I² = 0%; Analysis 1.3).

Subgroup analysis: follow‐up duration

There was no evidence of a subgroup difference between follow‐up durations (test for subgroup differences: Chi² = 0.15, df = 1 (P = 0.70), I² = 0%; Analysis 1.4).

1.4. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 4: Recurrence rate (Subgroup analysis according to follow‐up duration)

1.4. Wound infection

Based on 13 studies with 1568 participants (Akca 2005; Bukhari 2020; Daglar 2004; Dass 2012; Enshaei 2014; Enshaei 2020; Ertan 2005; Galal 2012; Khan 2013; Maghsudi 2020; Muzi 2010; Sakr 2006; Shabbir 2014), off‐midline closure probably results in lower rates of wound infection compared to the conventional midline closure procedure (RR 0.32, 95% CI 0.22 to 0.49; I² = 0%; 13 studies, 1568 participants; moderate‐certainty evidence; Analysis 1.5). We could not identify asymmetry in the funnel plot suggestive of publication bias (see Figure 3). Egger's test was not statistically significant (P = 0.8889).

1.5. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 5: Wound infection

3.

Funnel plot of comparison: 1 Off‐midline closure versus midline closure (Conventional), outcome: 1.6 Wound infection.

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 0.42, df = 3 (P = 0.94), I² = 0%; Analysis 1.5).

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 1.6).

1.6. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 6: Wound infection (Sensitivity analysis according to risk of bias)

1.5. Wound dehiscence

Based on 11 studies with 1389 participants (Abu 1999; Akca 2005; Dass 2012; Enshaei 2014; Ertan 2005; Galal 2012; Khan 2013; Maghsudi 2020; Muzi 2010; Sakr 2006; Shabbir 2014), off‐midline closure may result in lower rates of recurrence compared to the conventional midline closure procedure (RR 0.44, 95% CI 0.27 to 0.71; I² = 0%; 11 studies, 1389 participants; low‐certainty evidence; Analysis 1.7). The funnel plot was asymmetric, suggesting publication bias (see Figure 4). Egger's test was marginally statistically significant (P = 0.0090).

1.7. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 7: Wound dehiscence

4.

Funnel plot of comparison: 1 Off‐midline closure versus midline closure (Conventional), outcome: 1.8 Wound dehiscence.

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 1.41, df = 3 (P = 0.70), I² = 0%; Analysis 1.7).

1.6. Time to return to work

Based on six studies with 820 participants (Akca 2005; Dass 2012; Galal 2012; Maghsudi 2020; Muzi 2010; Shabbir 2014), off‐midline closure may reduce the time to wound healing compared to the conventional midline closure procedure (MD ‐3.72 days, 95% CI ‐6.11 to ‐1.33 days; I² = 99%; 6 studies, 820 participants; low‐certainty evidence; Analysis 1.8).

1.8. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 8: Time to return to work

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 0.01, df = 1 (P = 0.91), I² = 0%; Analysis 1.8).

1.7. Operative time

Based on six studies with 921 participants (Akca 2005; Dass 2012; Galal 2012; Maghsudi 2020; Muzi 2010; Sakr 2006), we found that participants who received off‐midline closure may have experienced longer operative times compared to those who received the conventional midline closure procedure (MD 15.43 minutes (min), 95% CI 7.56 to 23.30 min; I² = 99%; 6 studies, 921 participants; Analysis 1.9).

1.9. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 9: Operative time

1.8. Length of hospital stay

Based on 10 studies with 1279 participants (Akca 2005; Bukhari 2020; Dass 2012; Enshaei 2014; Enshaei 2020; Ertan 2005; Galal 2012; Muzi 2010; Sakr 2006; Shabbir 2014), we found that off‐midline closure was associated with a slightly shorter mean length of hospital stay compared to conventional midline closure (MD ‐0.57 days, 95% CI ‐1.01 to ‐0.14 days; I² = 95%; 10 studies, 1279 participants; Analysis 1.10). We abstained from constructing a funnel plot to investigate possible small study biases, given that the data from Akca 2005 and Muzi 2010 were not integrated into the meta‐analysis.

1.10. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 10: Length of hospital stay

It is important to note that the difference in mean length of hospital stay between off‐midline closure and conventional midline closure was only ‐0.57 days, which may not be clinically meaningful for individual patients. However, this small difference could potentially translate to significant cost savings for healthcare systems if applied to a large population. It is also important to consider that the minimal clinically important difference (MCID) for length of hospital stay in pilonidal sinus surgery has not been established. Therefore, while the reported result indicates a statistically significant difference between the two closure methods, the clinical relevance of this difference remains uncertain.

1.9. Postoperative or post‐procedural pain

Based on two studies with 320 participants (Akca 2005; Galal 2012), we found that participants who received off‐midline closure may have experienced less pain on the first postoperative day compared to those who received the conventional midline closure procedure (MD ‐2.53, 95% CI ‐3.60 to ‐1.45; I² = 95%; 2 studies, 320 participants; Analysis 1.11).

1.11. Analysis.

Comparison 1: Off‐midline closure versus midline closure (conventional), Outcome 11: Postoperative or post‐procedural pain

1.10. Quality of life

Ertan 2005 assessed quality of life using the SF‐36 questionnaire. Although the mean scores for each of the eight SF‐36 domains were reported (Table 5), summary scores for physical and mental components were not provided.

2. Ertan 2005: SF‐36 Quality of life scores at 12 months (Mean ± SD).

Item	Off‐midline (Limberg flap) N=50	Midline closure (Conventional) N=50	P value
General health perception	78.2 ±14.1	71.1 ±11.7	0.008
Physical functioning	77.6 ±14.8	74.8 ±12.9	NS
Social functioning	110.4 ±33.8	87.3 ±32.8	0.001
Role limitation ‐ physical	49.5 ±41.1	42.5 ±29.1	NS
Role limitation ‐ emotional	54.6 ±44.5	57.3 ±33.0	NS
Pain	67.5 ±18.4	54.5 ±14.0	0.001
Vitality	72.9 ±17.2	77.5 ±16.8	NS
Mental health	77.6 ±13.1	79.5 ±12.3	NS

NS = non‐significant

1.11. Cost‐effectiveness

None of the included studies reported data for this outcome.

2. Off‐midline closure versus tension‐free midline closure

Six studies with 559 analysed participants compared off‐midline closure to midline closure (tension‐free procedure) (Arnous 2019; Hemmingsson 2022; Okuş 2012; Rashidian 2014; Sevinc 2016; Tavassoli 2011). See: Table 2.

2.1. Time to wound healing

Based on one study with 116 participants (Hemmingsson 2022), off‐midline closure (Karydakis flap) may reduce the time to wound healing compared to the tension‐free procedure (median 14 days (95% CI 12 to 20 days) in the off‐midline closure group versus 51 days (95% CI 34 to 66 days) in the tension‐free midline closure group; P = 0.001; low‐certainty evidence).

Two other studies offer additional insights:

• Rashidian 2014 found that wound healing time for the off‐midline closure with the Limberg flap was slightly quicker at 1.71 weeks, compared to 1.92 weeks for the tension‐free midline closure.

• Tavassoli 2011 reported that suture removal – which served as a measure of wound healing in this work – took place on postoperative day 15.3 for participants who had an off‐midline closure with the Limberg flap, compared to postoperative day 20.08 for those who had a tension‐free midline closure (P = 0.08).

2.2. Proportion of wounds healed

Data were not available for the 1‐month and 6‐month time points but were reported at three and 12 months.

Proportion of wounds healed at three months

Based on one study with 115 participants (Hemmingsson 2022), off‐midline closure may increase wound healing rates at three months when compared to the conventional midline closure procedure (RR 1.24, 95% CI 1.06 to 1.46; 1 study, 115 participants; low‐certainty evidence; Analysis 2.1).

2.1. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 1: Proportion of wounds healed at 3 months

Proportion of wounds healed at 12 months

Based on one study with 115 participants (Hemmingsson 2022), off‐midline closure may result in little to no difference in wound healing rates at 12 months when compared to the conventional midline closure procedure (RR 1.00, 95% CI 0.95 to 1.05; 1 study, 115 participants; Analysis 2.2).

2.2. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 2: Proportion of wounds healed at 12 months

2.3. Recurrence rate

Based on six studies with 551 participants assessed at a wide range of follow‐up points (from 11 days to nearly 14 years) (Arnous 2019; Hemmingsson 2022; Okuş 2012; Rashidian 2014; Sevinc 2016; Tavassoli 2011), off‐midline closure may result in little to no difference in rates of recurrence compared to the tension‐free procedure (RR 0.69, 95% CI 0.30 to 1.61; I² = 14%; 6 studies, 551 participants; very low‐certainty evidence; Analysis 2.3).

2.3. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 3: Recurrence rate

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (Chi² = 2.15, df = 2 (P = 0.34), I² = 7.0%; Analysis 2.3).

Subgroup analysis: follow‐up duration

There was no evidence of a subgroup difference between follow‐up durations (test for subgroup differences: Chi² = 0.98, df = 1 (P = 0.32), I² = 0%; Analysis 2.4).

2.4. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 4: Recurrence rate (Subgroup analysis according to follow‐up duration)

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 2.5).

2.5. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 5: Recurrence rate (Sensitivity analysis according to risk of bias)

Sensitivity analysis: characteristics of tension‐free midline closure

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies using tension‐free midline closure that did not provide explicit descriptions of the distance from the edge of the fascia (Analysis 2.6).

2.6. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 6: Recurrence rate (Sensitivity analysis according to characteristics of tension‐free midline closure)

2.4. Wound infection

Based on six studies with 559 participants (Arnous 2019; Hemmingsson 2022; Okuş 2012; Rashidian 2014; Sevinc 2016; Tavassoli 2011), off‐midline closure (all methods used for this outcome involved the Limberg flap technique) may result in little to no difference in rates of wound infection compared to the tension‐free procedure (RR 0.44, 95% CI 0.16 to 1.17; I² = 0%; 6 studies, 559 participants; very low‐certainty evidence; Analysis 2.7).

2.7. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 7: Wound infection

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 2.14, df = 2 (P = 0.34), I² = 6.5%; Analysis 2.7).

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 2.8).

2.8. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 8: Wound infection (Sensitivity analysis according to risk of bias)

Sensitivity analysis: characteristics of tension‐free midline closure

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies using tension‐free midline closure that did not provide explicit descriptions of the distance from the edge of the fascia (Analysis 2.9).

2.9. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 9: Wound infection (Sensitivity analysis according to characteristics of tension‐free midline closure)

2.5. Wound dehiscence

Based on three studies with 250 participants (Arnous 2019; Rashidian 2014; Sevinc 2016), off‐midline closure may result in little to no difference in rates of wound dehiscence compared to the tension‐free procedure (RR 0.82, 95% CI 0.17 to 3.84; I² = 0%; 3 studies, 250 participants; very low‐certainty evidence; Analysis 2.10).

2.10. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 10: Wound dehiscence

Subgroup analysis: type of off‐midline closure

There was no evidence of a subgroup difference between types of off‐midline closure (test for subgroup differences: Chi² = 0.19, df = 1 (P = 0.66), I² = 0%; Analysis 2.10).

2.6. Time to return to work

Based on one study with 60 participants (Arnous 2019), off‐midline closure may result in increased time to return to work compared to tension‐free midline closure (MD 3.00 days, 95% CI 1.52 to 4.48 days; 1 study, 60 participants; low‐certainty evidence; Analysis 2.11).

2.11. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 11: Time to return to work

In Rashidian 2014, the average time to return to work in the off‐midline closure group was 4.2 days, and in the midline closure group it was five days.

2.7. Operative time

Based on three studies with 310 participants (Arnous 2019; Sevinc 2016; Tavassoli 2011), off‐midline closure may result in little to no difference in operative time compared to those who received the tension‐free procedure (MD 8.93 min, 95% CI ‐0.14 to 18.01 min; I² = 98%; 3 studies, 310 participants; Analysis 2.12).

2.12. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 12: Operative time

The Hemmingsson 2022 study offers additional insights: off‐midline closure (Karydakis flap) may increase the operative time compared to the tension‐free procedure (median 60 min (95% CI 50 to 69 min) in the off‐midline closure group versus 31.5 min (95% CI 25 to 36 min) in the tension‐free midline closure group; P＜0.001).

2.8. Length of hospital stay

Based on one study with 60 participants (Arnous 2019), we found that off‐midline closure was associated with a slightly longer mean length of hospital stay compared to tension‐free midline closure (MD 0.3 days, 95% CI ‐0.00 to 0.60 days; 1 study, 60 participants; Analysis 2.13).

2.13. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 13: Length of hospital stay

In Rashidian 2014, the average length of hospital stay in the off‐midline closure group was 1.35 (range 1 to 4) days and in the midline closure group it was 0.95 (range 0 to 1) days.

2.9. Postoperative or post‐procedural pain

Based on two studies with 160 participants (Arnous 2019; Tavassoli 2011), off‐midline closure may result in little to no difference in postoperative or post‐procedural pain on the first postoperative day compared to the tension‐free procedure (MD ‐0.69, 95% CI ‐2.65 to 1.27; I² = 97%; 2 studies, 160 participants; Analysis 2.14).

2.14. Analysis.

Comparison 2: Off‐midline closure versus midline closure (tension‐free), Outcome 14: Postoperative or post‐procedural pain

2.10. Quality of life

None of the included studies reported data for this outcome.

2.11. Cost‐effectiveness

None of the included studies reported data for this outcome.

3. Off‐midline closure (Karydakis flap) versus off‐midline closure (Limberg flap)

Nine studies with 899 analysed participants compared the Karydakis to the Limberg flap (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Kohla 2016; Riaz 2019; Sevinc 2016). See: Table 3.

3.1. Time to wound healing

Based on six studies with 526 participants (Alvandipour 2019; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Kohla 2016), the Karydakis flap probably results in little to no difference in time to wound healing compared to the Limberg flap (MD 0.36 days, 95% CI ‐1.49 to 2.22 days; I² = 64%; 6 studies, 526 participants; moderate‐certainty evidence; Analysis 3.1).

3.1. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 1: Time to wound healing

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 3.2).

3.2. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 2: Time to wound healing (Sensitivity analysis according to risk of bias)

3.2. Proportion of wounds healed

Data were not available for the 1‐month, 6‐month, and 12‐month time points.

Based on one study with 30 participants with follow‐up time not reported (Kohla 2016), the Karydakis flap may result in little to no difference to wound healing rate compared to the Limberg flap (RR 1.20, 95% CI 0.77 to 1.86; 1 study, 30 participants; low‐certainty evidence; Analysis 3.3).

3.3. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 3: Proportion of wounds healed

3.3. Recurrence rate

Based on nine studies with 890 participants with six to 41 months of follow‐up (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Kohla 2016; Riaz 2019; Sevinc 2016), the Karydakis flap may result in little to no difference in recurrence rate compared to the Limberg flap (RR 1.14, 95% CI 0.61 to 2.14; I² = 6%; 9 studies, 890 participants; low‐certainty evidence; Analysis 3.4).

3.4. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 4: Recurrence rate

Subgroup analysis: follow‐up duration

There was no evidence of a subgroup difference between follow‐up durations (test for subgroup differences: Chi² = 0.45, df = 1 (P = 0.50), I² = 0%; Analysis 3.4).

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 3.5).

3.5. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 5: Recurrence rate (Sensitivity analysis according to risk of bias)

3.4. Wound infection

Based on eight studies with 869 participants (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Riaz 2019; Sevinc 2016), the Karydakis flap may result in little to no difference in wound infection compared to the Limberg flap (RR 1.55, 95% CI 0.90 to 2.68; I² = 1%; 8 studies, 869 participants; low‐certainty evidence; Analysis 3.6).

3.6. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 6: Wound infection

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 3.7).

3.7. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 7: Wound infection (Sensitivity analysis according to risk of bias)

3.5. Wound dehiscence

Based on seven studies with 776 participants (Alvandipour 2019; Arslan 2014; Bali 2015; El Hadidi 2019; Kohla 2016; Riaz 2019; Sevinc 2016), the Karydakis flap may result in little to no difference in wound dehiscence compared to the Limberg flap (RR 1.20, 95% CI 0.41 to 3.50; I² = 56%; 7 studies, 776 participants; low‐certainty evidence; Analysis 3.8).

3.8. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 8: Wound dehiscence

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 3.9).

3.9. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 9: Wound dehiscence (Sensitivity analysis according to risk of bias)

3.6. Time to return to work

Based on six studies with 541 participants (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Kohla 2016), the Karydakis flap may result in little to no difference in time to return to work compared to the Limberg flap (MD ‐0.23 days, 95% CI ‐5.53 to 5.08 days; I² = 99%; 6 studies, 541 participants; low‐certainty evidence; Analysis 3.10).

3.10. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 10: Time to return to work

Sensitivity analysis: overall high risk of bias

The results of sensitivity analysis indicated that the pooled result was not substantially affected by the inclusion of studies with an overall high risk of bias (Analysis 3.11).

3.11. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 11: Time to return to work (Sensitivity analysis according to risk of bias)

3.7. Operative time

Based on nine studies with 899 participants (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; El Hadidi 2019; Elhiny 2021; Kohla 2016; Riaz 2019; Sevinc 2016), the Karydakis flap may result in a slightly shorter operative time compared to the Limberg flap (MD ‐7.38 min, 95% CI ‐10.10 to ‐4.67 min; I² = 89%; 9 studies, 899 participants; Analysis 3.12).

3.12. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 12: Operative time

3.8. Length of hospital stay

Based on five studies with 445 participants (Alvandipour 2019; Anwar 2021; Arslan 2014; Bali 2015; Elhiny 2021), the Karydakis flap may result in little to no difference in length of hospital stay compared to the Limberg flap (MD 0.25 days, 95% CI ‐0.40 to 0.90 days; I² = 97%; 5 studies, 445 participants; Analysis 3.13).

3.13. Analysis.

Comparison 3: Off‐midline (Karydakis flap) versus off‐midline (Limberg flap), Outcome 13: Length of hospital stay

Outcomes with no data

None of the studies for this comparison reported on:

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

4. Off‐midline closure (Bascom’s cleft lift) versus off‐midline closure (Limberg flap)

One study with 122 analysed participants compared Bascom’s cleft lift to the Limberg flap (Guner 2013).

4.1. Time to wound healing

Guner 2013 did not report this outcome in a way that allowed us to generate a summary statistic. Guner 2013 reported that Bascom’s cleft lift (mean 11.50 days, range 10 to 36 days) may result in little to no difference (P = 0.51) in time to wound healing compared to the Limberg flap (mean 11.55 days, range 10 to 23 days).

4.2. Recurrence rate

Based on one study with 122 participants and a median 13‐month follow‐up (Guner 2013), Bascom’s cleft lift may result in little to no difference in recurrence rate compared to the Limberg flap (RR 0.33, 95% CI 0.01 to 8.03; 1 study, 122 participants; Analysis 4.1).

4.1. Analysis.

Comparison 4: Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap), Outcome 1: Recurrence rate

4.3. Wound infection

Based on one study with 122 participants (Guner 2013), Bascom’s cleft lift may result in little to no difference in wound infection compared to the Limberg flap (RR 0.83, 95% CI 0.27 to 2.59; 1 study, 122 participants; Analysis 4.2).

4.2. Analysis.

Comparison 4: Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap), Outcome 2: Wound infection

4.4. Wound dehiscence

Based on one study with 122 participants (Guner 2013), Bascom’s cleft lift may result in little to no difference in wound dehiscence compared to the Limberg flap (RR 1.00, 95% CI 0.21 to 4.76; 1 study, 122 participants; Analysis 4.3).

4.3. Analysis.

Comparison 4: Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap), Outcome 3: Wound dehiscence

4.5. Operative time

Guner 2013 did not report this outcome in a way that allowed us to generate a summary statistic. Guner 2013 reported that Bascom’s cleft lift (mean 29 min, range 20 to 42 min) may result in a slightly shorter operative time (P < 0.0001) compared to the Limberg flap (mean 36 min, range 22 to 50 min).

4.6. Length of hospital stay

Guner 2013 did not report this outcome in a way that allowed us to generate a summary statistic. Guner 2013 reported that Bascom’s cleft lift (mean 1.28 days, range 1 to 3 days) may result in little to no difference (P = 0.28) in length of hospital stay compared to the Limberg flap (mean 1.43 days, range 1 to 7 days).

4.7. Quality of life

Guner 2013 did not report this outcome in a way that allowed us to generate a summary statistic. Guner 2013 assessed quality of life using the SF‐36 questionnaire. Although the mean scores for each of the eight SF‐36 domains were reported (Table 6), summary scores for physical and mental components were not provided.

3. Guner 2013: SF‐36 Quality of life scores at 1 month (Mean ± SD).

Item	Off‐midline (Limberg flap) N = 61	Off‐midline (Bascom cleft lift) N = 61	P value
General health	84.87 ± 22.29	89.79 ± 10.79	0.88
Physical functioning	92.29 ± 10.31	94.67 ± 8.74	0.18
Social functioning	80.12 ± 18.87	83.4 ± 13.64	0.54
Role limitation ‐ physical	76.64 ± 35.02	83.61 ± 26.17	0.41
Role limitation ‐ emotional	78.69 ± 24.37	83.07 ± 20.74	0.37
Pain	83.87 ± 14.48	84.06 ± 21.66	0.18
Vitality	78.74 ± 13.61	74.02 ± 17.95	0.18
Mental health	79.15 ± 15.36	77.11 ± 13.95	0.14

Outcomes with no data

Guner 2013 did not report on:

• proportion of wounds healed;

• time to return to work;

• postoperative or post‐procedural pain;

• cost‐effectiveness.

5. Off‐midline closure (modified Limberg flap) versus off‐midline closure (Limberg flap)

One study with 204 analysed participants compared the modified Limberg flap to the Limberg flap (Arslan 2014).

5.1. Time to wound healing

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in time to wound healing compared to the Limberg flap (MD 1.90 days, 95% CI ‐0.95 to 4.75 days; 1 study, 204 participants; Analysis 5.1).

5.1. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 1: Time to wound healing

5.2. Recurrence rate

Based on one study with 204 participants and a mean follow‐up of more than 30 months (Arslan 2014), the modified Limberg flap may result in little to no difference in recurrence rate compared to the Limberg flap (RR 0.30, 95% CI 0.06 to 1.43; 1 study, 204 participants; Analysis 5.2).

5.2. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 2: Recurrence rate

5.3. Wound infection

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in wound infection compared to the Limberg flap (RR 2.22, 95% CI 0.44 to 11.19; 1 study, 204 participants; Analysis 5.3).

5.3. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 3: Wound infection

5.4. Wound dehiscence

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in wound dehiscence compared to the Limberg flap (RR 1.78, 95% CI 0.33 to 9.49; 1 study, 204 participants; Analysis 5.4).

5.4. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 4: Wound dehiscence

5.5. Time to return to work

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in time to return to work compared to the Limberg flap (MD ‐1.00 days, 95% CI ‐2.56 to 0.56 days; 1 study, 204 participants; Analysis 5.5).

5.5. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 5: Time to return to work

5.6. Operative time

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in operative time compared to the Limberg flap (MD 1.80 min, 95% CI ‐0.14 to 3.74 min; 1 study, 204 participants; Analysis 5.6).

5.6. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 6: Operative time

5.7. Length of hospital stay

Based on one study with 204 participants (Arslan 2014), the modified Limberg flap may result in little to no difference in length of hospital stay compared to the Limberg flap (MD 0.00 days, 95% CI ‐0.13 to 0.13 days; 1 study, 204 participants; Analysis 5.7).

5.7. Analysis.

Comparison 5: Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap), Outcome 7: Length of hospital stay

Outcomes with no data

Arslan 2014 did not report on:

• proportion of wounds healed;

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

6. Off‐midline closure (modified Limberg flap) versus off‐midline closure (Karydakis flap)

Two studies with 344 analysed participants compared the modified Limberg flap to the Karydakis flap (Arslan 2014; Can 2010).

6.1. Time to wound healing

Based on two studies with 336 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in time to wound healing compared to the Karydakis flap (MD 2.12 days, 95% CI ‐2.19 to 6.43 days; I² = 72%; 2 studies, 336 participants; Analysis 6.1).

6.1. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 1: Time to wound healing

6.2. Recurrence rate

Based on two studies with 336 participants and a mean follow‐up of 16.8 to 34.5 months (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in recurrence rate compared to the Karydakis flap (RR 0.44, 95% CI 0.07 to 2.91; I² = 69%; 2 studies, 336 participants; Analysis 6.2).

6.2. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 2: Recurrence rate

6.3. Wound infection

Based on two studies with 344 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in wound infection compared to the Karydakis flap (RR 0.86, 95% CI 0.35 to 2.12; I² = 0%; 2 studies, 344 participants; Analysis 6.3).

6.3. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 3: Wound infection

6.4. Wound dehiscence

Based on two studies with 344 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in wound dehiscence compared to the Karydakis flap (RR 0.41, 95% CI 0.12 to 1.41; I² = 0%; 2 studies, 344 participants; Analysis 6.4).

6.4. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 4: Wound dehiscence

6.5. Time to return to work

Based on two studies with 336 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in time to return to work compared to the Karydakis flap (MD 1.30 days, 95% CI ‐0.50 to 3.10 days; I² = 44%; 2 studies, 336 participants; Analysis 6.5).

6.5. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 5: Time to return to work

6.6. Operative time

Based on two studies with 336 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in operative time compared to the Karydakis flap (MD 7.09 min, 95% CI ‐3.10 to 17.28 min; I² = 96%; 2 studies, 336 participants; Analysis 6.6).

6.6. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 6: Operative time

6.7. Length of hospital stay

Based on two studies with 336 participants (Arslan 2014; Can 2010), the modified Limberg flap may result in little to no difference in length of hospital stay compared to the Karydakis flap (MD ‐0.25 days, 95% CI ‐0.90 to 0.41 days; I² = 69%; 2 studies, 336 participants; Analysis 6.7).

6.7. Analysis.

Comparison 6: Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap), Outcome 7: Length of hospital stay

Outcomes with no data

Neither of the two studies for this comparison reported on:

• proportion of wounds healed;

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

7. Off‐midline closure (modified Karydakis flap) versus off‐midline closure (Karydakis flap)

One study with 100 analysed participants compared the modified Karydakis flap to the Karydakis flap (Kartal 2019).

7.1. Recurrence rate

Based on one study with 100 participants (Kartal 2019), the mean follow‐up period was determined to be 25 months, with a range of 18 to 33 months, across both groups. Throughout this follow‐up period, there were no observed recurrences in any of the patients from either group.

7.2. Wound infection

Based on one study with 100 participants (Kartal 2019), the modified Karydakis flap may result in little to no difference in wound infection compared to the Karydakis flap (RR 0.50, 95% CI 0.10 to 2.61; 1 study, 100 participants; Analysis 7.1).

7.1. Analysis.

Comparison 7: Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap), Outcome 1: Wound infection

7.3. Wound dehiscence

Based on one study with 100 participants (Kartal 2019), the modified Karydakis flap may result in little to no difference in wound dehiscence compared to the Karydakis flap (RR 0.40, 95% CI 0.08 to 1.97; 1 study, 100 participants; Analysis 7.2).

7.2. Analysis.

Comparison 7: Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap), Outcome 2: Wound dehiscence

7.4. Operative time

Based on one study with 100 participants (Kartal 2019), the modified Karydakis flap may result in a longer operative time compared to the Karydakis flap (MD 4.31 min, 95% CI 1.68 to 6.94 min; 1 study, 100 participants; Analysis 7.3).

7.3. Analysis.

Comparison 7: Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap), Outcome 3: Operative time

Outcomes with no data

Kartal 2019 did not report on:

• time to wound healing;

• proportion of wounds healed;

• time to return to work;

• length of hospital stay;

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

8. Off‐midline closure (modified Limberg flap) versus off‐midline closure (modified Karydakis flap)

One study with 120 analysed participants compared the modified Limberg flap to the modified Karydakis flap (Bessa 2013).

8.1. Recurrence rate

Based on a single study involving 120 participants and a median follow‐up of 20.5 months (Bessa 2013), the modified Limberg flap may result in little to no difference in recurrence rate compared to the modified Karydakis flap (RR 2.00, 95% CI 0.19 to 21.47; 1 study, 120 participants; Analysis 8.1).

8.1. Analysis.

Comparison 8: Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap), Outcome 1: Recurrence rate

8.2. Wound infection

Based on a single study involving 120 participants (Bessa 2013), the modified Limberg flap may result in little to no difference in wound infection compared to the modified Karydakis flap (RR 1.50, 95% CI 0.26 to 8.66; 1 study, 120 participants; Analysis 8.2).

8.2. Analysis.

Comparison 8: Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap), Outcome 2: Wound infection

8.3. Wound dehiscence

Based on a single study involving 120 participants (Bessa 2013), the modified Limberg flap may result in little to no difference in wound dehiscence compared to the modified Karydakis flap (RR 1.83, 95% CI 0.72 to 4.64; 1 study, 120 participants; Analysis 8.3).

8.3. Analysis.

Comparison 8: Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap), Outcome 3: Wound dehiscence

8.4. Operative time

Based on a single study involving 120 participants (Bessa 2013), the modified Limberg flap may result in a longer operative time compared to the modified Karydakis flap (MD 19.31 min, 95% CI 18.30 to 20.32 min; 1 study, 120 participants; Analysis 8.4).

8.4. Analysis.

Comparison 8: Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap), Outcome 4: Operative time

Outcomes with no data

Bessa 2013 did not report on:

• time to wound healing;

• proportion of wounds healed;

• time to return to work;

• length of hospital stay;

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

9. Off‐midline closure (modified Limberg flap) versus off‐midline closure (Burow’s triangle advancement flap)

One study with 100 analysed participants compared the modified Limberg flap to Burow’s triangle advancement flap (Saydam 2015).

9.1. Recurrence rate

Based on one study with 100 participants and a mean follow‐up of 12 months (Saydam 2015), the modified Limberg flap may result in a higher, but uncertain, recurrence rate compared to Burow's triangle advancement flap, with more people experiencing recurrence in the modified Limberg flap (RR 3.00, 95% CI 0.13 to 71.92; 1 study, 100 participants; Analysis 9.1).

9.1. Analysis.

Comparison 9: Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap), Outcome 1: Recurrence rate

9.2. Wound infection

Based on one study with 100 participants (Saydam 2015), it is uncertain whether the modified Limberg flap prevents wound infection (RR 5.00, 95% CI 0.25 to 101.58; 1 study, 100 participants; Analysis 9.2).

9.2. Analysis.

Comparison 9: Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap), Outcome 2: Wound infection

9.3. Wound dehiscence

Based on one study with 100 participants (Saydam 2015), the modified Limberg flap may result in little to no difference in wound dehiscence rate compared to Burow's triangle advancement flap (RR 1.50, 95% CI 0.26 to 8.60; 1 study, 100 participants; Analysis 9.3).

9.3. Analysis.

Comparison 9: Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap), Outcome 3: Wound dehiscence

9.4. Operative time

Based on one study with 100 participants (Saydam 2015), the modified Limberg flap may result in little to no difference in operative time compared to Burow’s triangle advancement flap (MD ‐1.01 min, 95% CI ‐4.63 to 2.61 min; 1 study, 100 participants; Analysis 9.4).

9.4. Analysis.

Comparison 9: Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap), Outcome 4: Operative time

9.5. Length of hospital stay

Based on one study with 100 participants (Saydam 2015), the modified Limberg flap may result in a slightly shorter hospital stay compared to Burow’s triangle advancement flap (MD ‐1.09 days, 95% CI ‐1.51 to ‐0.67 days; I² = 69%; 1 study, 100 participants; Analysis 9.5).

9.5. Analysis.

Comparison 9: Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap), Outcome 5: Length of hospital stay

Outcomes with no data

Saydam 2015 did not report on:

• time to wound healing;

• proportion of wounds healed;

• time to return to work;

• postoperative or post‐procedural pain;

• quality of life;

• cost‐effectiveness.

Discussion

Summary of main results

We included 33 studies with 3667 analysed participants. The median or average age of participants in these studies varied from 21.0 to 34.2 years; most were male. Geographically, most trials were based in the Middle East, with Turkey, Egypt, and Iran being the most represented countries.

Off‐midline closure versus conventional midline closure

We found a total of 14 studies with 1614 participants that compared off‐midline closure with conventional midline closure. Off‐midline closure probably reduced the time to wound healing, resulted in lower rates of recurrence, wound infection, and wound dehiscence, and reduced the time to return to work compared to conventional midline closure. However, the certainty of evidence for these outcomes was mostly low to moderate. There may be little to no difference in the proportion of wounds healed between the two techniques, with very low‐certainty evidence.

For the other outcomes, off‐midline closure was associated with longer operative time and shorter length of hospital stay, although the latter may not be clinically meaningful. Participants who received off‐midline closure may have experienced less postoperative or post‐procedural pain. We found that no available data could contribute to our meta‐analyses regarding quality of life or cost‐effectiveness outcomes.

Off‐midline closure versus tension‐free midline closure

Six studies with 559 analysed participants compared off‐midline closure with tension‐free midline closure. The main outcomes revealed that off‐midline closure may reduce time to wound healing, increase wound healing rates at three months, and increase time to return to work with low‐certainty evidence, while little to no difference was observed in recurrence rate, wound infection, and wound dehiscence with very low‐certainty evidence.

Other outcomes showed that off‐midline closure may result in little to no difference in operative time and postoperative or post‐procedural pain, and may lead to a slightly longer length of hospital stay, though the clinical relevance of this difference is uncertain. No data were available for the outcomes of quality of life and cost‐effectiveness.

Off‐midline (Karydakis flap) versus off‐midline (Limberg flap) closure

Nine studies with 890 analysed participants compared the off‐midline Karydakis flap to the Limberg flap for wound closure after surgical treatment for pilonidal sinus. The Karydakis flap resulted in little to no difference in time to wound healing, proportion of wounds healed, recurrence rate, wound infection, wound dehiscence, and time to return to work when compared to the Limberg flap, with low‐ to moderate‐certainty evidence for these outcomes.

In terms of other outcomes, the results showed that the Karydakis flap may result in shorter operative time and little to no difference in length of hospital stay compared to the Limberg flap. None of the nine studies for this comparison reported on postoperative or post‐procedural pain, quality of life, or cost‐effectiveness.

Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap) closure

One study with 122 analysed participants compared Bascom’s cleft lift versus the Limberg flap. The results showed that Bascom’s cleft lift may result in little to no difference in time to wound healing compared to the Limberg flap. The results showed that Bascom’s cleft lift may result in little to no difference in recurrence rate, wound infection, and wound dehiscence compared to the Limberg flap. The results also showed that Bascom’s cleft lift may result in little to no difference in operative time and length of hospital stay compared to the Limberg flap.

The single study for this comparison did not report on proportion of wounds healed, time to return to work, postoperative or post‐procedural pain, or cost‐effectiveness outcomes. Data on quality of life were insufficiently reported for analysis.

Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap) closure

One study with 204 participants compared the modified Limberg flap with the Limberg flap. The results of the study suggest that there may be little to no difference between the two techniques with respect to time to wound healing, recurrence rate, wound infection, wound dehiscence, and time to return to work. The results also suggest that there may be little to no difference between the two techniques for operative time and length of hospital stay.

The single study for this comparison did not report on proportion of wounds healed, postoperative or post‐procedural pain, quality of life, and cost‐effectiveness.

Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap) closure

Two studies with 344 participants compared the modified Limberg flap to the Karydakis flap. Regarding the most important outcomes, the evidence suggests that the modified Limberg flap and Karydakis flap may result in little to no difference in time to wound healing, recurrence rate, wound infection, wound dehiscence, and time to return to work. Regarding the rest of the outcomes, no differences were observed in operative time and length of hospital stay between the two procedures.

Neither study for this comparison reported the proportion of wounds healed, postoperative or post‐procedural pain, quality of life, and cost‐effectiveness.

Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap) closure

One study with 100 participants compared the modified Karydakis flap with the Karydakis flap. The results of the study suggest that there may be little to no difference between the two techniques with respect to wound infection and wound dehiscence. Longer operative time was observed for the modified Karydakis flap group.

Throughout the follow‐up period of this study, there were no observed recurrences in any of the patients from either group. This study did not report on time to wound healing, proportion of wounds healed, time to return to work, length of hospital stay, postoperative or post‐procedural pain, quality of life, and cost‐effectiveness.

Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap) closure

One study with 120 participants compared the modified Limberg flap with the modified Karydakis flap. The results suggest that there may be little to no difference between the two techniques with respect to recurrence rate, wound infection, and wound dehiscence. The study found that the modified Limberg flap may result in longer operative time compared to the modified Karydakis flap.

This study did not report on time to wound healing, proportion of wounds healed, time to return to work, length of hospital stay, postoperative or post‐procedural pain, quality of life, and cost‐effectiveness.

Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap) closure

One study with 100 participants compared the modified Limberg flap with Burow’s triangle advancement flap. There is insufficient evidence to determine whether the modified Limberg flap has a different recurrence, wound infection, or wound dehiscence rate when compared to Burow's triangle advancement flap. The results suggest that the modified Limberg flap may result in little to no difference in operative time and slightly shorter length of hospital stay compared to Burow’s triangle advancement flap.

This study did not report on time to wound healing, proportion of wounds healed, time to return to work, postoperative or post‐procedural pain, quality of life, or cost‐effectiveness.

Overall completeness and applicability of evidence

The studies we included focused on participants with either new onset or recurrent pilonidal sinus disease. We excluded studies involving participants who presented with an abscess. Due to the frequently unclear reporting on clinical subgroups, we did not attempt subgroup or sensitivity analyses based on clinical variables such as disease severity or co‐interventions.

Blinding participants, surgeons, and outcome assessors in surgical RCTs often presents challenges, and in some situations, it is downright unfeasible. Thus, we recommend interpreting the findings from this surgical review cautiously. The potential bias arising from the absence of blinding could influence outcomes.

Moreover, a significant portion of the trials came from the Middle East, and most were from single‐centre institutions. This geographical concentration introduces two potential biases.

• Geographic bias: the results may not be generalisable to other populations, as they primarily represent the practices and outcomes of Middle Eastern institutions. Healthcare systems, surgical techniques, and patient demographics in these regions may differ from those in other parts of the world.

• Institutional bias: since most studies in the current review come from single‐centre institutions, the results may be influenced by the specific practices and expertise of those institutions. This could limit the generalisability of the findings to other centres with different resources, patient populations, or surgical expertise.

In relation to tension‐free midline closure, surgeons continue to confront the challenge of standardizing the degree of subcutaneous tissue release. The studies examined demonstrate variability in the extent of subcutaneous tissue release employed.

Whilst our protocol did not impose age restrictions on participants (Cai 2022), it is crucial to highlight that most studies predominantly involved adults (see Table 4). Even though pilonidal sinus treatments in children often mirror those in adults, a pertinent question remains: can outcomes observed in adults reliably be extrapolated to children (Hardy 2019)? The literature hints at potential differences between these demographic groups. For instance, findings suggest that paediatric and adolescent patients may present distinct recurrence patterns (Hardy 2019; Maasewerd 2023), indicating a possible inclination towards earlier recurrences in this age group, irrespective of the treatment method. To navigate the path from paediatric to adult treatment strategies with confidence, a deeper understanding of the driving factors behind recurrent episodes in paediatric pilonidal sinus disease is indispensable. However, research targeting the paediatric population is noticeably limited (Grabowski 2019). This gap leaves unaddressed the specific concerns and challenges inherent to this younger demographic. According to a systematic review, some non‐randomised studies have indicated that some treatment approaches, including off‐midline primary closure, marsupialisation, and minimally invasive techniques, are not only comparable in their outcomes but also consistently deliver better results than other treatments (Hardy 2019). These outcomes particularly pertain to recurrence, wound complications, and healing times in children. However, the methodological quality of these non‐randomised studies focusing on children often falls short (Hardy 2019). Common shortcomings included small sample sizes, limited follow‐up periods, and a tangible risk of bias, which casts doubts on their findings.

Certainty of the evidence

The main limitations of the body of evidence were the following.

• Study limitations

  • Prospective registration and protocols: most studies were not prospectively registered and did not have published protocols or analysis plans, making it difficult to assess the risk of selective reporting bias.

  • Random sequence generation and allocation concealment: most of the included studies did not provide sufficient details about these methodological aspects, raising concerns about potential selection bias.

  • Sample size calculations: the majority of the studies did not present clear sample size calculations or failed to achieve the required sample size, which may reduce the statistical power to detect significant differences between interventions.

  • Outcome assessment and time points: a majority of studies neglected to detail outcome assessment procedures or specify the time points chosen for analysis.

  • Survival analysis: most studies did not employ survival analysis for time‐to‐event outcomes such as time to wound healing and recurrence, potentially limiting the accuracy and generalisability of the findings related to these outcomes.

  • CONSORT guideline adherence: when presenting their findings, numerous studies failed to adhere to the CONSORT guidelines (Butcher 2022; Schulz 2010), which may compromise the overall quality and transparency of the research.

• Inconsistency: there was considerable heterogeneity observed in some outcomes of our comparisons in this review. This inconsistency may stem from several factors, as follows.

  • Differences in outcome measure definitions: variability in the definitions used for outcome measures across studies can contribute to the observed heterogeneity, making it challenging to draw definitive conclusions from the pooled results.

  • Variations in surgical approaches: the extent of tissue excision, incision depth, suture material selection, and suturing techniques employed varied across the included studies. These differences in surgical methods could lead to discrepancies in the observed outcomes and contribute to the overall heterogeneity.

• Indirectness: in the comparison of off‐midline closure versus midline closure (conventional or tension‐free procedure), the analysis included only a limited number of off‐midline techniques, such as the Limberg flap and modified Limberg flap. However, several other off‐midline closure techniques for pilonidal sinus disease exist, which were not included in this review because no studies investigated these techniques. This indirectness may affect the generalisability of the review's findings, as the conclusions may not be applicable to other off‐midline closure techniques that were not included in the analysis. Future research should aim to investigate and compare a broader range of off‐midline closure techniques to provide a more comprehensive understanding of their effectiveness in treating pilonidal sinus disease.

• Imprecision: due to the fact that most of the included studies did not meet the optimal information size (OIS) criteria, the results of most outcomes are affected by imprecision. This means that the available evidence may not be sufficient to provide conclusive findings, and the confidence intervals of the effect estimates may be wider than desired.

• Publication bias: due to the paucity of trials, we evaluated the risk of publication bias using funnel plots for only three outcomes pertaining to off‐midline closure versus midline closure (recurrence rate, wound infection, and wound dehiscence). We discovered evidence of asymmetry in wound dehiscence.

Potential biases in the review process

We categorised surgical techniques into midline and off‐midline closures. Within each category, variations in surgical approaches emerged, such as the extent of tissue excision, incision depth, degree of tissue releasing, suture material selection, and suturing technique employed. This heterogeneity constitutes a known limitation when aggregating data in meta‐analyses. However, embracing such compromises is crucial for generating meaningful comparisons (Al‐Khamis 2010).

Additionally, many studies failed to appropriately define surgical outcomes, and studies used different methods to assess similar outcomes, some of which seemed not to be the most appropriate or valid method (e.g. telephone interviewing or postal questionnaires to assess sinus recurrence). Future studies should define outcomes consistently and use standard assessment tools.

Lastly, participant dropout is an expected phenomenon in RCTs, particularly in long‐term studies that demand sustained participant engagement. This engagement can lead to increased non‐compliance, as seen in studies requiring participants to complete extensive questionnaires repeatedly. Consequently, an absence of missing data points might be considered unusual (Mol 2023). However, only a few of the included studies adequately addressed participant dropout. Intriguingly, several studies report a 100% follow‐up rate, even extending up to three years, a claim that warrants careful scrutiny. This oversight is significant, given the characteristics of the study population, which is a predominantly young, male, and often mobile population (Brown 2019).

Agreements and disagreements with other studies or reviews

Since 2014, several systematic reviews have been published addressing certain comparisons of midline and off‐midline wound closure methods following surgical treatment for pilonidal sinus (e.g. classical Limberg versus classical Karydakis flaps) (Bi 2020; Doll 2019; Emile 2021; Enriquez‐Navascues 2014; Gavriilidis 2019; Hardy 2019; Milone 2018; Prassas 2018; Ray 2020; Sahebally 2019; Stauffer 2018). These reviews exhibit various methodological limitations, such as the absence of a protocol outlining the review methods, insufficiently comprehensive searches, restrictive language limitations, and a lack of integration of the evidence quality with the results, as exemplified by not employing GRADE methods.

Our current review included all comparisons of midline and off‐midline wound closure methods following surgical treatment for pilonidal sinus. All of these other reviews identified studies that we also found, but the decisions to include or exclude specific studies varied between reviews. For instance, we excluded seven quasi‐randomised trials from our review (Ates 2011; Cihan 2006; Ersoy 2009; Karim 2020; Kayal 2014; Tokac 2015; Zorlu 2016). The Enriquez‐Navascues 2014 review classified Muzi 2010 in the tension‐free midline closure group, which differed from our approach. Our team assessed the surgical details described in the article: "In group A, patients were placed in the prone position; the area to be excised was marked on the skin, and a rhomboid incision to the presacral fascia, including the sinus and its extensions, was made down. The cavity was then closed with absorbable sutures in a double layer with alternating deep and superficial sutures to minimize tissue traction and skin edge impairment. Thus, the skin edges were approximated without tension using interrupted sutures of absorbable 3/0 suture or liquid skin adhesive" (Muzi 2010). Although it was a type of modified midline closure method, it differs from the definition of "tension‐free midline closure" in the German National Guideline on the management of pilonidal disease (Iesalnieks 2019), and more closely resembles conventional midline closure. Therefore, we categorised this study with the conventional midline closure group.

Authors' conclusions

Implications for practice.

This Cochrane review examines the midline and off‐midline wound closure options for pilonidal sinus. Most evidence came from studies in young adults. The findings indicate moderate‐ to very low‐certainty evidence that there may be benefits associated with off‐midline flap surgical closure compared to conventional midline closure techniques. Most surgical outcomes demonstrated there may be evidence of a difference favouring off‐midline surgical closure. When comparing off‐midline flap closure to tension‐free midline closure, low‐ to very low‐certainty evidence shows that there may be reduced time to wound healing in off‐midline closure, and there may be little to no difference in recurrence rate, wound infection, and wound dehiscence.

Comparisons between different off‐midline flap techniques showed there may be no evidence of a difference in most outcomes, with moderate‐ to low‐certainty evidence. Evidence is currently uncertain regarding whether there is any advantage of one flap technique over the other for most of the evaluated outcomes. The choice of procedure is likely to be based on the clinician's preference, experience, patient characteristics, and the patient's preferences.

Implications for research.

Given the low‐ to moderate‐certainty evidence for most comparisons and outcomes included in this review, key considerations for future research include the following.

• Focus on paediatric and adolescent populations: the literature is markedly scant when it comes to addressing the unique concerns of paediatric and adolescent patients. To bridge this gap, meticulously executed, large‐scale randomised controlled trials (RCTs) are indispensable to pinpoint the optimal treatment modalities for pilonidal sinus in these younger populations.

• Prospective registration of trials: many RCTs on this topic were not registered or retrospectively registered. Ensuring prospective registration of clinical trials is crucial to maintain transparency and improve the quality of research.

• Ensuring adequate power in future pilonidal sinus surgery trials: many studies exhibit a lack of evidence for sample size calculations, or the details provided are vague, resulting in inadequate sample sizes. Trials should be sufficiently powered, potentially using sample size calculations and engaging multiple centres for recruitment.

• Standardisation of outcome measures: the multiplicity and disparity of outcome measures in pilonidal sinus surgery research, along with the lack of standardisation, have been extensively documented (Al‐Khamis 2010; Brown 2019; Characteristics of included studies). To address the absence of standardisation, developing core outcome sets for pilonidal sinus surgery research is essential.

• Robust assessment of outcomes: future trials should employ valid methods to evaluate surgical outcomes, such as time to wound healing and recurrence, analysed using survival methods. Few studies accounted for dropout, even though this is likely to be high given the young, mainly male, and often mobile population (Brown 2019). An emphasis on long‐term outcomes is also crucial.

• Inclusion of health economic assessments: cost‐effectiveness is rarely considered in current research. Future studies should ideally incorporate health economic evaluations, including the assessment of costs to primary and secondary care, as well as societal costs in terms of work absences and similar metrics.

• Adherence to CONSORT guidelines: trials should comply with the recommended standards outlined in the CONSORT guidelines to ensure transparency, methodological rigour, and reliable research.

History

Protocol first published: Issue 5, 2022

Risk of bias

Risk of bias for analysis 2.11 Time to return to work.

Study

Bias

Randomisation process

Deviations from intended interventions

Missing outcome data

Measurement of the outcome

Selection of the reported results

Overall

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Authors' judgement

Support for judgement

Arnous 2019

Low risk of bias

"Randomisation was achieved through computer software and the results were sealed into envelopes." No imbalances are apparent in the baseline characteristics among the randomized groups.

Some concerns

“The trial was single blinded as the patients gave consent to participate in the study, yet they were not aware of the type of procedure they will undergo. The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation.” The surgical procedure may not deviate significantly from the pre‐determined or intended strategy. Analyses excluding 2 eligible trial participants in each group post‐randomization. For continuous outcomes, availability of data from 95% of the participants will often be sufficient.

Low risk of bias

“Lost in follow‐up: 1/31 vs 1/31.”. For continuous outcomes, availability of data from 95% of the participants will often be sufficient. The proportion of missing outcome data is sufficiently low that any bias is too small to be of importance.

Some concerns

“Time away from work was defined as the date on which patient returned to normal activities including employment and leisure activities from the date of surgery." "The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation." However, outcome assessor is the participant, even if a blinded interviewer is questioning the participant and completing a questionnaire on their behalf. Participant‐reported outcomes. The outcome assessment is potentially influenced by knowledge of intervention received, leading to a judgement of at least ‘Some concerns’.

Some concerns

There was no pre‐registered protocol or statistical plan for this study. However, it is worth mentioning that the trial was retrospectively registered at www.researchregistry.com (researchregistry3601). The results were unlikely to have been selected from multiple eligible outcome measurements.

Some concerns

Based on judgements of previous domains.

Appendices

Appendix 1. Glossary of terms

Term	Definition
Bascom procedure	An off‐midline surgical procedure focusing on removal of the hairs and follicles with small wounds. This procedure consists of a lateral incision, removal of the hair, and excision of the sinus tract with small incisions overlying each sinus. The lateral incision outside the midline allows access to and excision of the subcutaneous diseased tissue.
Bascom cleft closure (cleft lift)	An off‐midline surgical procedure. This technique entails triangular excision of all diseased tissue and reconstruction of the defect with a flap outside the midline. This method completely eliminates the cleft.
Classical rhomboid (Limberg) flap	A transposition flap used to cover a rhomboid‐shaped defect after removing diseased tissue. This technique allows tension‐free closure of the rhomboid defect and donor site.
Karydakis technique	An off‐midline surgical procedure. The Karydakis technique refers to suturing mobilised fasciocutaneous tissue laterally to the sacrococcygeal fascia, thereby covering the defect left by midline‐involved excision.
Marsupialisation (partial closure)	After the pilonidal sinus is excised, the wound is partially closed to shape a permanently open pouch, which allows for easy drainage of the fluid. Marsupialisation provides the patient with a smaller wound compared with wounds that are left open to granulate.
Modified rhomboid (Limberg) flap	The rhomboid excision is asymmetric, with the lower pole of the flap positioned off‐midline, which is different from the classical rhomboid flap.
Purulent discharge	A pus‐containing fluid that is thick, opaque, and typically yellow‐white in colour. It is frequently associated with infection.
Serous discharge	A watery fluid that is thin, odourless, and clear to pale yellow in colour. It is mainly present in non‐infected or healing wounds.
V‐Y advancement flap	An off‐midline surgical procedure. The V‐Y advancement flap adopts the strategy of V‐fashion preparation and Y‐fashion closure.
Wound dehiscence	A partial or total separation of the edges of a surgical wound.
Z‐plasty	An off‐midline surgical procedure. Z‐plasty is a surgical technique that entails raising, transposing, and suturing two interdigitating triangular flaps to eliminate the gluteal cleft.

Appendix 2. Search strategies

Cochrane Wounds Specialised Register

1 MESH DESCRIPTOR Pilonidal Sinus EXPLODE ALL AND INREGISTER

2 pilonidal* AND INREGISTER

3 #1 OR #2

The Cochrane Central Register of Controlled Clinical Trials (CENTRAL)

#1 MeSH descriptor: [Pilonidal Sinus] explode all trees

#2 pilonidal*:ti,ab,kw

#3 #1 or #2

Ovid MEDLINE

1 exp Pilonidal Sinus/

2 pilonidal*.ti,ab,kw.

3 1 or 2

4 randomly.ab.

5 randomized controlled trial.pt.

6 controlled clinical trial.pt.

7 randomi?ed.ab.

8 placebo.ab.

9 clinical trials as topic.sh.

10 randomly.ab.

11 trial.ti.

12 or/5‐11

13 exp animals/ not humans.sh.

14 12 not 13

15 3 and 14

Ovid Embase

1 exp pilonidal sinus/

2 pilonidal*.ti,ab,kw.

3 1 or 2

4 Randomized controlled trial/

5 Controlled clinical study/

6 Random$.ti,ab.

7 randomization/

8 intermethod comparison/

9 placebo.ti,ab.

10 (compare or compared or comparison).ti.

11 ((evaluated or evaluate or evaluating or assessed or assess) and (compare or compared or comparing or comparison)).ab.

12 (open adj label).ti,ab.

13 ((double or single or doubly or singly) adj (blind or blinded or blindly)).ti,ab.

14 double blind procedure/

15 parallel group$1.ti,ab.

16 (crossover or cross over).ti,ab.

17 ((assign$ or match or matched or allocation) adj5 (alternate or group$1 orintervention$1 or patient$1 or subject$1 or participant$1)).ti,ab.

18 (assigned or allocated).ti,ab.

19 (controlled adj7 (study or design or trial)).ti,ab.

20 (volunteer or volunteers).ti,ab.

21 human experiment/

22 trial.ti.

23 or/4‐22

24 (random$ adj sampl$ adj7 (cross section$ or questionnaire$1 or survey$ or database$1)).ti,ab. not (comparative study/ or controlled study/ or randomi?ed controlled.ti,ab. or randomly assigned.ti,ab.)

25 Cross‐sectional study/ not (randomized controlled trial/ or controlled clinical study/ or controlled study/ or randomi?ed controlled.ti,ab. or control group$1.ti,ab.)

26 (((case adj control$) and random$) not randomi?ed controlled).ti,ab.

27 (Systematic review not (trial or study)).ti.

28 (nonrandom$ not random$).ti,ab.

29 Random field$.ti,ab.

30 (random cluster adj3 sampl$).ti,ab.

31 (review.ab. and review.pt.) not trial.ti.

32 we searched.ab. and (review.ti. or review.pt.)

33 update review.ab.

34 (databases adj4 searched).ab.

35 (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys or trout or marmoset$1).ti. and animal experiment/

36 Animal experiment/ not (human experiment/ or human/)

37 or/24‐36

38 23 not 37

39 3 and 38

EBSCO CINAHL Plus

S1 (MH "Pilonidal Cyst")

S2 TI pilonidal OR AB pilonidal

S3 S1 OR S2

S4 MH randomized controlled trials

S5 MH double‐blind studies

S6 MH single‐blind studies

S7 MH random assignment

S8 MH pretest‐posttest design

S9 MH cluster sample

S10 TI (randomised OR randomized)

S11 AB (random*)

S12 TI (trial)

S13 MH (sample size) AND AB (assigned OR allocated OR control)

S14 MH (placebos)

S15 PT (randomized controlled trial)

S16 AB (control W5 group)

S17 MH (crossover design) OR MH (comparative studies)

S18 AB (cluster W3 RCT)

S19 MH animals+

S20 MH (animal studies)

S21 TI (animal model*)

S22 S19 OR S20 OR S21

S23 MH (human)

S24 S22 NOT S23

S25 S4 OR S5 OR S6 OR S7 OR S8 OR S9 OR S10 OR S11 OR S12 OR S13 OR S14 OR S15 OR S16 OR S17 OR S18

S26 S25 NOT S24

S27 S3 AND S26

US National Institutes of Health Ongoing Trials Register (ClinicalTrials.gov)

Pilonidal Sinus

Pilonidal

World Health Organization International Clinical Trials Registry Platform

Pilonidal

Appendix 3. RoB 2: randomised trials

The response options for the (numbered) signalling questions are:

• Yes (Y);

• Probably yes (PY);

• Probably no (PN);

• No (N);

• No information (NI);

• Not applicable (NA).

The possible risk of bias judgements are:

• Low risk of bias;

• Some concerns;

• High risk of bias.

The response options for the optional questions are:

• Not applicable (NA);

• Favours experimental;

• Towards null;

• Away from null;

• Unpredictable.

Domain 1: Risk of bias arising from the randomisation process

1. Was the allocation sequence random?

2. Was the allocation sequence concealed until participants were enrolled and assigned to interventions?

3. Did baseline differences between intervention groups suggest a problem with the randomisation process?

• Risk of bias judgement

• Optional: What is the predicted direction of bias arising from the randomisation process? 

Domain 2: Risk of bias due to deviations from the intended interventions (effect of assignment to intervention)

1. Were participants aware of their assigned intervention during the trial?

2. Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?

3. If Y/PY/NI to 1 or 2: Were there deviations from the intended intervention that arose because of the trial context?

4. If Y/PY to 3: Were these deviations likely to have affected the outcome?

5. If Y/PY/NI to 4: Were these deviations from intended intervention balanced between groups?

6. Was an appropriate analysis used to estimate the effect of assignment to intervention?

7. If N/PN/NI to 6: Was there potential for a substantial impact (on the result) of the failure to analyse participants in the group to which they were randomised?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to deviations from intended interventions?

Domain 2: Risk of bias due to deviations from the intended interventions (effect of adhering to intervention)

1. Were participants aware of their assigned intervention during the trial?

2. Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?

3. [If applicable:] If Y/PY/NI to 1 or 2: Were important non‐protocol interventions balanced across intervention groups?

4. [If applicable:] Were there failures in implementing the intervention that could have affected the outcome?

5. [If applicable:] Was there non‐adherence to the assigned intervention regimen that could have affected participants’ outcomes?

6. If N/PN/NI to 3, or Y/PY/NI to 4 or 5: Was an appropriate analysis used to estimate the effect of adhering to the intervention?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to deviations from intended interventions?

Domain 3: Missing outcome data

1. Were data for this outcome available for all, or nearly all, participants randomised?

2. If N/PN/NI to 1: Is there evidence that the result was not biased by missing outcome data?

3. If N/PN to 2: Could missingness in the outcome depend on its true value?

4. If Y/PY/NI to 3: Is it likely that missingness in the outcome depended on its true value?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to missing outcome data?

Domain 4: Risk of bias in measurement of the outcome

1. Was the method of measuring the outcome inappropriate?

2. Could measurement or ascertainment of the outcome have differed between intervention groups?

3. If N/PN/NI to 1 and 2: Were outcome assessors aware of the intervention received by study participants?

4. If Y/PY/NI to 3: Could assessment of the outcome have been influenced by knowledge of intervention received?

5. If Y/PY/NI to 4: Is it likely that assessment of the outcome was influenced by knowledge of intervention received?

• Risk of bias judgement

• Optional: What is the predicted direction of bias in measurement of the outcome?

Domain 5: Risk of bias in selection of the reported result

1. Were the data that produced this result analysed in accordance with a pre‐specified analysis plan that was finalised before unblinded outcome data were available for analysis?

2. Is the numerical result being assessed likely to have been selected, on the basis of the results, from...

   1. ...multiple eligible outcome measurements (e.g. scales, definitions, time points) within the outcome domain?

   2. ...multiple eligible analyses of the data?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to selection of the reported result?

Overall risk of bias

• Risk of bias judgement

• Optional: What is the overall predicted direction of bias for this outcome?)

Appendix 4. RoB 2: cluster‐randomised trials

The response options for the (numbered) signalling questions are:

• Yes (Y);

• Probably yes (PY);

• Probably no (PN);

• No (N);

• No information (NI);

• Not applicable (NA).

The possible risk of bias judgements are:

• Low risk of bias;

• Some concerns;

• High risk of bias.

The response options for the optional questions are:

• Not applicable (NA);

• Favours experimental;

• Towards null;

• Away from null;

• Unpredictable.

Domain 1a: Risk of bias arising from the randomisation process

1. Was the allocation sequence random?

2. Was the allocation sequence concealed until clusters were enrolled and assigned to interventions?

3. Did baseline differences between intervention groups suggest a problem with the randomisation process?

• Risk of bias judgement

• Optional: What is the predicted direction of bias arising from the randomisation process?

Domain 1b: Risk of bias arising from the timing of identification or recruitment of participants in a cluster‐randomised trial

1. Were all the individual participants identified and recruited (if appropriate) before randomisation of clusters?

2. If N/PN/NI to 1: Is it likely that selection of individual participants was affected by knowledge of the intervention assigned to the cluster?

3. Were there baseline imbalances that suggest differential identification or recruitment of individual participants between intervention groups?

• Risk of bias judgement

• Optional: What is the predicted direction of bias arising from the timing of identification and recruitment of participants?

Domain 2a: Risk of bias due to deviations from the intended interventions (effect of assignment to intervention)

1. 1. Were participants aware that they were in a trial?

   2. If Y/PY/NI to 1a: Were participants aware of their assigned intervention during the trial?

2. Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?

3. If Y/PY/NI to 1 or 2: Were there deviations from the intended intervention that arose because of the trial context?

4. If Y/PY to 3: Were these deviations likely to have affected the outcome?

5. If Y/PY/NI to 4: Were these deviations from intended intervention balanced between groups?

6. Was an appropriate analysis used to estimate the effect of assignment to intervention?

7. If N/PN/NI to 6: Was there potential for a substantial impact (on the result) of the failure to analyse participants in the group to which they were randomised?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to deviations from intended interventions? 

Domain 2b: Risk of bias due to deviations from the intended interventions (effect of adhering to intervention)

1. Were participants aware of their assigned intervention during the trial?

2. Were carers and people delivering the interventions aware of participants' assigned intervention during the trial?

3. [If applicable:] If Y/PY/NI to 1 or 2: Were important non‐protocol interventions balanced across intervention groups?

4. [If applicable:] Were there failures in implementing the intervention that could have affected the outcome?

5. [If applicable:] Was there non‐adherence to the assigned intervention regimen that could have affected participants’ outcomes?

6. If N/PN/NI to 3, or Y/PY/NI to 4 or 5: Was an appropriate analysis used to estimate the effect of adhering to the intervention?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to deviations from intended interventions?

Domain 3: Risk of bias due to missing outcome data

1. 1. Were data for this outcome available for all clusters that recruited participants?

   2. Were data for this outcome available for all, or nearly all, participants within clusters?

2. If N/PN/NI to 1a or 1b: Is there evidence that the result was not biased by missing data?

3. If N/PN to 2 Could missingness in the outcome depend on its true value?

4. If Y/PY/NI to 3: Is it likely that missingness in the outcome depended on its true value?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to missing outcome data?

Domain 4: Risk of bias in measurement of the outcome

1. Was the method of measuring the outcome inappropriate?

2. Could measurement or ascertainment of the outcome have differed between intervention groups?

   1. If N/PN/NI to 1 and 2: Were outcome assessors aware that a trial was taking place?

   2. 4.3b If Y/PY/NI to 3a: Were outcome assessors aware of the intervention received by study participants?

3. If Y/PY/NI to 3b: Could assessment of the outcome have been influenced by knowledge of intervention received?

4. If Y/PY/NI to 4: Is it likely that assessment of the outcome was influenced by knowledge of intervention received?

• Risk of bias judgement

• Optional: What is the predicted direction of bias in measurement of the outcome?

Domain 5: Risk of bias in selection of the reported result

1. Were the data that produced this result analysed in accordance with a pre‐specified analysis plan that was finalised before unblinded outcome data were available for analysis?

2. Is the numerical result being assessed likely to have been selected, on the basis of the results, from...

   1. ...multiple eligible outcome measurements (e.g. scales, definitions, time points) within the outcome domain?

   2. ...multiple eligible analyses of the data?

• Risk of bias judgement

• Optional: What is the predicted direction of bias due to selection of the reported result?

Overall risk of bias

• Risk of bias judgement

• Optional: What is the overall predicted direction of bias for this outcome?

Data and analyses

Comparison 1. Off‐midline closure versus midline closure (conventional).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Time to wound healing	3	300	Mean Difference (IV, Random, 95% CI)	‐5.23 [‐7.55, ‐2.92]
1.1.1 Limberg flap	2	180	Mean Difference (IV, Random, 95% CI)	‐5.32 [‐9.24, ‐1.40]
1.1.2 Modified Limberg flap	1	120	Mean Difference (IV, Random, 95% CI)	‐5.00 [‐6.19, ‐3.81]
1.2 Proportion of wounds healed	2	207	Risk Ratio (M‐H, Random, 95% CI)	1.13 [0.92, 1.39]
1.2.1 Limberg flap	1	46	Risk Ratio (M‐H, Random, 95% CI)	1.29 [1.02, 1.63]
1.2.2 Karydakis flap	1	161	Risk Ratio (M‐H, Random, 95% CI)	1.05 [0.97, 1.14]
1.3 Recurrence rate	13	1492	Risk Ratio (M‐H, Random, 95% CI)	0.22 [0.11, 0.45]
1.3.1 Limberg flap	8	975	Risk Ratio (M‐H, Random, 95% CI)	0.19 [0.06, 0.65]
1.3.2 Modified Limberg flap	2	180	Risk Ratio (M‐H, Random, 95% CI)	0.25 [0.05, 1.14]
1.3.3 Karydakis flap	1	161	Risk Ratio (M‐H, Random, 95% CI)	0.23 [0.05, 1.03]
1.3.4 Rotational flap	2	176	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.04, 3.21]
1.4 Recurrence rate (Subgroup analysis according to follow‐up duration)	13	1492	Risk Ratio (M‐H, Random, 95% CI)	0.22 [0.11, 0.45]
1.4.1 Follow‐up＜12 months	3	276	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.04, 3.21]
1.4.2 Follow‐up ≥ 12 months	10	1216	Risk Ratio (M‐H, Random, 95% CI)	0.21 [0.10, 0.45]
1.5 Wound infection	13	1568	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.22, 0.49]
1.5.1 Limberg flap	8	1051	Risk Ratio (M‐H, Random, 95% CI)	0.31 [0.19, 0.50]
1.5.2 Modified Limberg flap	2	180	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.10, 0.83]
1.5.3 Karydakis flap	1	161	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.08, 2.08]
1.5.4 Rotational flap	2	176	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.11, 2.05]
1.6 Wound infection (Sensitivity analysis according to risk of bias)	12	1308	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.21, 0.51]
1.6.1 Limberg flap	7	791	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.19, 0.54]
1.6.2 Modified Limberg flap	2	180	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.10, 0.83]
1.6.3 Karydakis flap	1	161	Risk Ratio (M‐H, Random, 95% CI)	0.42 [0.08, 2.08]
1.6.4 Rotational flap	2	176	Risk Ratio (M‐H, Random, 95% CI)	0.47 [0.11, 2.05]
1.7 Wound dehiscence	11	1389	Risk Ratio (M‐H, Random, 95% CI)	0.44 [0.27, 0.71]
1.7.1 Limberg flap	8	1028	Risk Ratio (M‐H, Random, 95% CI)	0.45 [0.26, 0.78]
1.7.2 Modified Limberg flap	1	120	Risk Ratio (M‐H, Random, 95% CI)	0.25 [0.03, 2.17]
1.7.3 Karydakis flap	1	161	Risk Ratio (M‐H, Random, 95% CI)	0.52 [0.05, 5.61]
1.7.4 Rotational flap	1	80	Risk Ratio (M‐H, Random, 95% CI)	0.09 [0.01, 1.59]
1.8 Time to return to work	6	820	Mean Difference (IV, Random, 95% CI)	‐3.72 [‐6.11, ‐1.33]
1.8.1 Limberg flap	4	640	Mean Difference (IV, Random, 95% CI)	‐3.65 [‐6.78, ‐0.52]
1.8.2 Modified Limberg flap	2	180	Mean Difference (IV, Random, 95% CI)	‐3.89 [‐6.96, ‐0.83]
1.9 Operative time	6	921	Mean Difference (IV, Random, 95% CI)	15.43 [7.56, 23.30]
1.9.1 Limberg flap	4	640	Mean Difference (IV, Random, 95% CI)	18.27 [9.05, 27.49]
1.9.2 Modified Limberg flap	1	120	Mean Difference (IV, Random, 95% CI)	14.60 [12.02, 17.18]
1.9.3 Karydakis flap	1	161	Mean Difference (IV, Random, 95% CI)	5.10 [2.98, 7.22]
1.10 Length of hospital stay	10	1279	Mean Difference (IV, Random, 95% CI)	‐0.57 [‐1.01, ‐0.14]
1.10.1 Limberg flap	5	762	Mean Difference (IV, Random, 95% CI)	‐0.93 [‐1.41, ‐0.45]
1.10.2 Modified Limberg flap	2	180	Mean Difference (IV, Random, 95% CI)	‐1.43 [‐1.95, ‐0.92]
1.10.3 Karydakis flap	1	161	Mean Difference (IV, Random, 95% CI)	0.50 [0.36, 0.64]
1.10.4 Rotational flap	2	176	Mean Difference (IV, Random, 95% CI)	0.05 [‐0.04, 0.15]
1.11 Postoperative or post‐procedural pain	2	320	Mean Difference (IV, Random, 95% CI)	‐2.53 [‐3.60, ‐1.45]
1.11.1 Limberg flap	1	200	Mean Difference (IV, Random, 95% CI)	‐2.00 [‐2.12, ‐1.88]
1.11.2 Modified Limberg flap	1	120	Mean Difference (IV, Random, 95% CI)	‐3.10 [‐3.57, ‐2.63]

Comparison 2. Off‐midline closure versus midline closure (tension‐free).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
2.1 Proportion of wounds healed at 3 months	1	115	Risk Ratio (M‐H, Random, 95% CI)	1.24 [1.06, 1.46]
2.2 Proportion of wounds healed at 12 months	1	115	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.95, 1.05]
2.3 Recurrence rate	6	551	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.30, 1.61]
2.3.1 Limberg flap	4	293	Risk Ratio (M‐H, Random, 95% CI)	0.34 [0.09, 1.33]
2.3.2 Karydakis flap	1	108	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.27, 2.91]
2.3.3 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.31, 7.17]
2.4 Recurrence rate (Subgroup analysis according to follow‐up duration)	6	551	Risk Ratio (M‐H, Random, 95% CI)	0.69 [0.30, 1.61]
2.4.1 Follow‐up ＜12months	1	100	Risk Ratio (M‐H, Random, 95% CI)	0.25 [0.03, 2.16]
2.4.2 Follow‐up ≥ 12months	5	451	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.33, 2.03]
2.5 Recurrence rate (Sensitivity analysis according to risk of bias)	5	491	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.39, 1.89]
2.5.1 Limberg flap	3	233	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.12, 2.14]
2.5.2 Karydakis flap	1	108	Risk Ratio (M‐H, Random, 95% CI)	0.89 [0.27, 2.91]
2.5.3 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.31, 7.17]
2.6 Recurrence rate (Sensitivity analysis according to characteristics of tension‐free midline closure)	2	243	Risk Ratio (M‐H, Random, 95% CI)	1.22 [0.36, 4.11]
2.6.1 Limberg flap	1	93	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.13, 6.11]
2.6.2 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.31, 7.17]
2.7 Wound infection	6	559	Risk Ratio (M‐H, Random, 95% CI)	0.44 [0.16, 1.17]
2.7.1 Limberg flap	4	293	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.12, 1.29]
2.7.2 Karydakis flap	1	116	Risk Ratio (M‐H, Random, 95% CI)	0.10 [0.01, 1.89]
2.7.3 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.16, 14.06]
2.8 Wound infection (Sensitivity analysis according to risk of bias)	5	499	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.19, 1.80]
2.8.1 Limberg flap	3	233	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.14, 2.60]
2.8.2 Karydakis flap	1	116	Risk Ratio (M‐H, Random, 95% CI)	0.10 [0.01, 1.89]
2.8.3 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.16, 14.06]
2.9 Wound infection (Sensitivity analysis according to characteristics of tension‐free midline closure)	2	243	Risk Ratio (M‐H, Random, 95% CI)	1.22 [0.22, 6.92]
2.9.1 Limberg flap	1	93	Risk Ratio (M‐H, Random, 95% CI)	0.90 [0.06, 13.93]
2.9.2 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.16, 14.06]
2.10 Wound dehiscence	3	250	Risk Ratio (M‐H, Random, 95% CI)	0.82 [0.17, 3.84]
2.10.1 Limberg flap	2	100	Risk Ratio (M‐H, Random, 95% CI)	0.67 [0.11, 3.96]
2.10.2 Limberg flap & Karydakis flap	1	150	Risk Ratio (M‐H, Random, 95% CI)	1.51 [0.06, 36.53]
2.11 Time to return to work	1	60	Mean Difference (IV, Random, 95% CI)	3.00 [1.52, 4.48]
2.12 Operative time	3	310	Mean Difference (IV, Random, 95% CI)	8.93 [‐0.14, 18.01]
2.12.1 Limberg Flap	2	160	Mean Difference (IV, Random, 95% CI)	12.68 [‐1.53, 26.89]
2.12.2 Limberg flap & Karydakis flap	1	150	Mean Difference (IV, Random, 95% CI)	1.54 [‐0.98, 4.05]
2.13 Length of hospital stay	1	60	Mean Difference (IV, Random, 95% CI)	0.30 [‐0.00, 0.60]
2.14 Postoperative or post‐procedural pain	2	160	Mean Difference (IV, Random, 95% CI)	‐0.69 [‐2.65, 1.27]

Comparison 3. Off‐midline (Karydakis flap) versus off‐midline (Limberg flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
3.1 Time to wound healing	6	526	Mean Difference (IV, Random, 95% CI)	0.36 [‐1.49, 2.22]
3.2 Time to wound healing (Sensitivity analysis according to risk of bias)	5	339	Mean Difference (IV, Random, 95% CI)	0.02 [‐2.23, 2.26]
3.3 Proportion of wounds healed	1	30	Risk Ratio (M‐H, Random, 95% CI)	1.20 [0.77, 1.86]
3.4 Recurrence rate	9	890	Risk Ratio (M‐H, Random, 95% CI)	1.14 [0.61, 2.14]
3.4.1 Follow‐up ＜12 months	4	524	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.49, 2.10]
3.4.2 Follow‐up ≥12 months	5	366	Risk Ratio (M‐H, Random, 95% CI)	1.84 [0.38, 8.92]
3.5 Recurrence rate (Sensitivity analysis according to risk of bias)	8	712	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.41, 2.04]
3.5.1 Follow‐up ＜12 months	2	268	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.23, 1.74]
3.5.2 Follow‐up ≥12 months	6	444	Risk Ratio (M‐H, Random, 95% CI)	1.36 [0.36, 5.17]
3.6 Wound infection	8	869	Risk Ratio (M‐H, Random, 95% CI)	1.55 [0.90, 2.68]
3.7 Wound infection (Sensitivity analysis according to risk of bias)	7	682	Risk Ratio (M‐H, Random, 95% CI)	1.42 [0.78, 2.56]
3.8 Wound dehiscence	7	776	Risk Ratio (M‐H, Random, 95% CI)	1.20 [0.41, 3.50]
3.9 Wound dehiscence (Sensitivity analysis according to risk of bias)	6	589	Risk Ratio (M‐H, Random, 95% CI)	0.59 [0.30, 1.15]
3.10 Time to return to work	6	541	Mean Difference (IV, Random, 95% CI)	‐0.23 [‐5.53, 5.08]
3.11 Time to return to work (Sensitivity analysis according to risk of bias)	5	354	Mean Difference (IV, Random, 95% CI)	0.06 [‐6.06, 6.18]
3.12 Operative time	9	899	Mean Difference (IV, Random, 95% CI)	‐7.38 [‐10.10, ‐4.67]
3.13 Length of hospital stay	5	445	Mean Difference (IV, Random, 95% CI)	0.25 [‐0.40, 0.90]

Comparison 4. Off‐midline (Bascom’s cleft lift) versus off‐midline (Limberg flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
4.1 Recurrence rate	1	122	Risk Ratio (M‐H, Random, 95% CI)	0.33 [0.01, 8.03]
4.2 Wound infection	1	122	Risk Ratio (M‐H, Random, 95% CI)	0.83 [0.27, 2.59]
4.3 Wound dehiscence	1	122	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.21, 4.76]

Comparison 5. Off‐midline (modified Limberg flap) versus off‐midline (Limberg flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
5.1 Time to wound healing	1	204	Mean Difference (IV, Random, 95% CI)	1.90 [‐0.95, 4.75]
5.2 Recurrence rate	1	204	Risk Ratio (M‐H, Random, 95% CI)	0.30 [0.06, 1.43]
5.3 Wound infection	1	204	Risk Ratio (M‐H, Random, 95% CI)	2.22 [0.44, 11.19]
5.4 Wound dehiscence	1	204	Risk Ratio (M‐H, Random, 95% CI)	1.78 [0.33, 9.49]
5.5 Time to return to work	1	204	Mean Difference (IV, Random, 95% CI)	‐1.00 [‐2.56, 0.56]
5.6 Operative time	1	204	Mean Difference (IV, Random, 95% CI)	1.80 [‐0.14, 3.74]
5.7 Length of hospital stay	1	204	Mean Difference (IV, Random, 95% CI)	0.00 [‐0.13, 0.13]

Comparison 6. Off‐midline (modified Limberg flap) versus off‐midline (Karydakis flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
6.1 Time to wound healing	2	336	Mean Difference (IV, Random, 95% CI)	2.12 [‐2.19, 6.43]
6.2 Recurrence rate	2	336	Risk Ratio (M‐H, Random, 95% CI)	0.44 [0.07, 2.91]
6.3 Wound infection	2	344	Risk Ratio (M‐H, Random, 95% CI)	0.86 [0.35, 2.12]
6.4 Wound dehiscence	2	344	Risk Ratio (M‐H, Random, 95% CI)	0.41 [0.12, 1.41]
6.5 Time to return to work	2	336	Mean Difference (IV, Random, 95% CI)	1.30 [‐0.50, 3.10]
6.6 Operative time	2	344	Mean Difference (IV, Random, 95% CI)	7.09 [‐3.10, 17.28]
6.7 Length of hospital stay	2	344	Mean Difference (IV, Random, 95% CI)	‐0.25 [‐0.90, 0.41]

Comparison 7. Off‐midline (modified Karydakis flap) versus off‐midline (Karydakis flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
7.1 Wound infection	1	100	Risk Ratio (M‐H, Random, 95% CI)	0.50 [0.10, 2.61]
7.2 Wound dehiscence	1	100	Risk Ratio (M‐H, Random, 95% CI)	0.40 [0.08, 1.97]
7.3 Operative time	1	100	Mean Difference (IV, Random, 95% CI)	4.31 [1.68, 6.94]

Comparison 8. Off‐midline (modified Limberg flap) versus off‐midline (modified Karydakis flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
8.1 Recurrence rate	1	120	Risk Ratio (M‐H, Random, 95% CI)	2.00 [0.19, 21.47]
8.2 Wound infection	1	120	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.26, 8.66]
8.3 Wound dehiscence	1	120	Risk Ratio (M‐H, Random, 95% CI)	1.83 [0.72, 4.64]
8.4 Operative time	1	120	Mean Difference (IV, Random, 95% CI)	19.31 [18.30, 20.32]

Comparison 9. Off‐midline (modified Limberg flap) versus off‐midline (Burow’s triangle advancement flap).

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
9.1 Recurrence rate	1	100	Risk Ratio (M‐H, Random, 95% CI)	3.00 [0.13, 71.92]
9.2 Wound infection	1	100	Risk Ratio (M‐H, Random, 95% CI)	5.00 [0.25, 101.58]
9.3 Wound dehiscence	1	100	Risk Ratio (M‐H, Random, 95% CI)	1.50 [0.26, 8.60]
9.4 Operative time	1	100	Mean Difference (IV, Random, 95% CI)	‐1.01 [‐4.63, 2.61]
9.5 Length of hospital stay	1	100	Mean Difference (IV, Random, 95% CI)	‐1.09 [‐1.51, ‐0.67]

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Abu 1999.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: A system of sequentially numbered sealed envelopes Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: United Arab Emirates Setting: Hospital Sample size: 46 Number randomised: 24 to intervention (Limberg flap) 22 to comparator (Conventional midline closure) Number analysed: 24 to intervention (Limberg flap) 22 to comparator (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 23 years (range 17 to 30); Control group: 21 years (range 18 to 34). Male (n, %): Intervention group: 19, 79.2%; Control group: 18, 81.8%. BMI (mean ± SD): Not mentioned. Type of presentation Primary disease Intervention group: 24 Control group: 22 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Adults with symptomatic sinuses of the natal cleft. Exclusion criteria for the study: Participants with acute abscesses or recurrent sinuses.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drains: Yes. Methylene blue: Yes. Perioperative antibiotic: Yes. Given a single dose of 1.5 g intravenously with the induction of anaesthesia.
Outcomes	Outcomes of interest in the review: Primary outcomes: Proportion of wounds healed (10th postoperative day) Recurrence rate Secondary outcomes: Adverse effects (wound infection (data not available), wound dehiscence) Time to return to work (data not available) Length of hospital stay (data not available) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Follow‐up: The mean follow‐up for both groups was 18 months.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Akca 2005.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A computer‐generated table of random numbers. The randomisation was carried out by an independent computer consultant. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 200 Number randomised: 100 to intervention, 100 to comparator 100 to intervention (Limberg flap) 100 to comparator (Conventional midline closure) Number analysed: 100 to intervention, 100 to comparator 100 to intervention (Limberg flap) 100 to comparator (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 26 (range 15 to 60); Control group: 28 (range 17 to 43). Male (n, %): Intervention group: 85, 85.0%; Control group: 83, 83.0%. BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 0 Control group: 0 Recurrent disease Intervention group: 0 Control group: 0 Unspecified (abscess (14 per cent)) Intervention group: 100 Control group: 100 Inclusion criteria for the study: Adults with symptomatic sinuses of the natal cleft. Exclusion criteria for the study: Patients who had undergone previous surgery for SPD (17 patients) and those who did not accept randomisation (five patients) were excluded.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes; Methylene blue: No; Perioperative antibiotic: Yes. A single dose of antibiotic prophylaxis was used immediately before incision.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence). Length of hospital stay (days) (median, IQR) Operative time (minutes) (median, IQR) Time to return to work (days) (median, IQR) Postoperative or post‐procedural pain (VAS, 1st postoperative day) (median, IQR) Other outcomes reported by the study: Time to first mobilisation Other adverse effects Follow‐up: Intervention group: median 28 (range 23 to 36) months Comparator group: median 29 (range 23 to 38) months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Alvandipour 2019.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: "According to the following formula from a study by Bali et al. (Bali 2015), and consider a drop‐out rate of 23% a sample size of 26 subjects per group provided approximately a power of 0.9 for comparing the 2 treatment groups" Randomisation method: Computer‐based table of randomisation Blinding: A single‐blind study for participants CONSORT flow diagram: Yes
Participants	Location: Iran Setting: Hospital Sample size: 65 Number randomised: 37 to intervention (Karydakis flap) 28 to comparator (Limberg flap) Number analysed: 37 to intervention (Karydakis flap) 27 to comparator (Limberg flap). One participant was lost during follow‐up due to family problems. Baseline characteristics: Age (years, mean ± SD): Intervention group: 25.89 ± 9.48; Control group: 34.19 ± 10.47 (P = 0.002). Male (n, %): Intervention group: 21, 56.8%; Control group: 9, 33.3% (P = 0.064). BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 37 Control group: 28 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with intergluteal pilonidal disease who had been referred to Imam Khomeini Hospital and whose ages ranged from 15 to 65 years. Exclusion criteria for the study: (1) Elderly patients with comorbid diseases, (2) scars from a previous pilonidal surgery, (3) abscess presentations, (4) diabetes mellitus, (5) immunodeficiency, (6) neurological disorders, (7) drug addiction or alcoholism, (8) American Society of Anesthesiologists physical status classification grade III–IV, (9) age < 15 years or > 65 years, and (10) orifice located more than 3 cm from the sinus centre.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. The patients on the operating table were administered antibiotic prophylaxis using intravenous cefazolin (1 g).
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) Operative time (minutes) Time to return to work (days) Other outcomes reported by the study: Postoperative or post‐procedural pain (time point was not reported) Other adverse effects Follow‐up: 6 months
Funding sources	Not mentioned
Declarations of interest	No potential conflict of interest relevant to this article was reported.
Notes	Publication status: Published

Anwar 2021.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: "The sample size was calculated with G power 3.9 at 5% level of significance and 90% power of the test which came 68 with 34 patients in each group." Randomisation method: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Pakistan Setting: Hospital Sample size: 70 Number randomised: 35 to intervention (Karydakis flap) 35 to comparator (Limberg flap) Number analysed: 35 to intervention (Karydakis flap) 34 to comparator (Limberg flap) One participant was lost during follow‐up Baseline characteristics: Age (years, mean ± SD): Intervention group: 26.54 ± 5.98 Control group: 25.97 ± 6.03 Male (n, %): Intervention group: 32, 94.1% Control group: 30, 88.2% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 35 Control group: 35 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Participants presented to the hospital with pilonidal sinus disease. Exclusion criteria for the study: Those with recurrent disease, uncontrolled diabetes, coagulopathy, or acute pilonidal abscess were excluded from the study.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. The procedures were performed under spinal anaesthesia with a single dose of 3rd generation cephalosporin given to all patients prior to skin incision.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) Operative time (minutes) Time to return to work Other outcomes reported by the study: Cost Postoperative or post‐procedural pain (time point was not reported) Other adverse effects Follow‐up: All patients were followed for six months.
Funding sources	Not mentioned
Declarations of interest	None declared
Notes	Publication status: Published

Arnous 2019.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: "The sample size was calculated based on the analysis of the primary outcome (recurrence of sacrococcygeal pilonidal disease at 12 months of follow‐up). Estimating that the recurrence rate after excision and primary midline closure can reach up to 30% and recurrence after Limberg flap can reach up to 3% according to the published literature, a sample size of 56 patients equally divided on both groups was estimated to be necessary to have a study power of 80% with alpha level set at 5%. To compensate for loss to follow‐up and dropout, a sample size of 60 patients was finally included in the study." Randomisation method: Randomisation was achieved through computer software and the results were sealed into envelopes. Blinding: "The trial was single blinded as the patients gave consent to participate in the study, yet they were not aware of the type of procedure they will undergo. The assessment of the outcome of both procedures was done by a surgical resident who was unaware of the nature of the study and group allocation." CONSORT flow diagram: Yes
Participants	Location: Egypt Setting: Hospital Sample size: 62 Number randomised: 31 to intervention (1 participant was lost), 31 to comparator (1 participant was lost) 31 to intervention (Limberg flap) 31 to comparator (Tension‐free midline closure) Number analysed: 30 to intervention, 30 to comparator 30 to intervention (Limberg flap) Lost to follow‐up (n = 1) 30 to comparator (Tension‐free midline closure) Lost to follow‐up (n = 1) Baseline characteristics: Age (years, mean ± SD): Intervention group: 23.8 ± 5.9 Control group: 25.1 ± 7.4 Male (n, %): Intervention group: 27 male, 3 female; control group: 25 male, 5 female Intervention group: 27, 90.0% Control group: 25, 83.3% BMI (mean ± SD): Intervention group: 26.5 ± 4.4 Control group: 27.4 ± 3.9 Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 30 Control group: 30 Inclusion criteria for the study: Adult patients of both sexes with primary or recurrent sacrococcygeal pilonidal disease. Exclusion criteria for the study: Acute presentation as pilonidal abscess Accidentally discovered sacrococcygeal pilonidal disease with no reported symptoms Pregnancy Psychiatric disease preventing surgical intervention Declined to participate in the study Unfit for anaesthesia
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Patients received antibiotics in the form of third‐generation cephalosporin before the incision.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) Operative time (minutes) Postoperative or post‐procedural pain (VAS, 1st postoperative day) Time to return to work Other outcomes reported by the study: Other adverse effects Time for drain removal Postoperative or post‐procedural pain on the fourth postoperative day, and at suture removal Follow‐up: Intervention group: mean 16.4 ± 3.7 months Control group: mean 17.03 ± 4 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published In the CONSORT flow diagram, the category 'Lost to follow‐up (give reasons)' listed one participant (n = 1); however, the trial did not provide any reason for this loss to follow‐up.

Arslan 2014.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: "The reported rate of postoperative complications associated with flap procedures varies between 1 and 13 %, thus it was determined that 112 patients per treatment arm would allow for the detection of a 10 % difference in the rate of complications with 80 % power and 5 % confidence." Randomisation method: Block randomisation Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 330 Number randomised: 110 to Modified Limberg flap 110 to Karydakis flap 110 to Limberg flap Number analysed: 108 to modified Limberg flap 2 participant was lost 91 to Karydakis flap 19 participant was lost 96 to Limberg flap 14 participant was lost Baseline characteristics: Age (years, mean ± SD): Modified Limberg flap: 24.7 ± 5.1 Karydakis flap: 24.7 ± 5.1 Limberg flap: 26.5 ± 5.9 Male (n, %): Modified Limberg flap: 95, 88.0% Karydakis flap: 77, 84.6% Limberg flap: 68, 70.8% BMI (mean ± SD): Not mentioned Acute abscesses: "If the patients admitted to our hospital with the diagnosis of sacrococcygeal PSD had abscesses, the abscesses were drained first and the patients were operated on at least 10 days later; those with active infection were operated on after antibiotic treatment." Type of presentation Primary disease 110 to modified Limberg flap: (2 participant was lost) 110 to Karydakis flap: (19 participant was lost) 110 to Limberg flap: (14 participant was lost) Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients scheduled for surgical intervention because of pilonidal sinus disease. Exclusion criteria for the study: Recurrent cases, ASA group higher than III, obesity (body mass index > 35 kg/m2), insulin‐dependent diabetes, drug and alcohol addiction.
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Off‐midline (Karydakis flap) Off‐midline (Limberg flap) Use of additional interventions in all groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. 1 g cefazolin sodium was intravenously administered while the patients were on the operating table.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) Operative time (minutes) Time to return to work (days) Other outcomes reported by the study: Other adverse effects Duration of drainage (days) Time until recurrence (months) Patient satisfaction (at 12 months) Follow‐up: Follow‐up visits took place on day 14 and at 1, 3, 6, and 12 months after surgery. In addition, patients were followed up by phone calls at 6‐month intervals (minimum 24, maximum 41 months). Modified Limberg flap: 32.9 ± 4.6 months Karydakis flap: 33.3 ± 5 months Limberg flap: 34.5 ± 5.3 months
Funding sources	Not mentioned
Declarations of interest	None
Notes	Publication status: Published

Bali 2015.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 71 Number randomised: 34 to intervention group (Karydakis flap) 37 to control group (Limberg flap) Number analysed: 34 to intervention group (Karydakis flap) 37 to control group (Limberg flap) Baseline characteristics: Age (years, mean ± SD): Intervention group: 23.5 Control group: 25 Male (n, %): Intervention group: 32, 94.1% Control group: 32, 86.5% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 0 Control group: 0 Recurrent disease Intervention group: 34 Control group: 37 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients willing to give written informed consent. Presence of recurrent pilonidal sinus. Adults (over 16 years of age) undergoing surgery for recurrent pilonidal sinus. International normalising ratio (INR) of less than 1.5. Prothrombin time (PT) of less than 15 s. Partial thromboplastin (PTT) time near normal. Platelet count greater than 50,000 per mm3 to limit the risk of bleeding. No infection at time of surgery. Exclusion criteria for the study: Patients not willing to give informed consent. Age less than 16 years. Patient presenting with different conditions mimicking pilonidal sinus. Chronic medical conditions, such as diabetes mellitus, renal failure, or immunosuppression.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Yes Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (median, range) Length of hospital stay (days) (median, range) Operative time (minutes) (median, range) Other outcomes reported by the study: Painless toilet seating (days) Patient satisfaction Postoperative or post‐procedural pain (median, range) (time point was not reported) Other adverse effects Time to first mobilisation Follow‐up: Intervention group: mean 27.6 ± 8.32 months Control group: mean 28.2 ± 7.65 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Bessa 2013.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: The authors used the mean operative times for the original Karydakis technique compared with the modified Limberg technique from the Can 2010 study to calculate the effect size d using Cohen’s formula. This initial power calculation suggested that a minimum of 56 patients per group would be required to detect a true difference of this size with a power of 80% at the 5% level of significance. Randomisation method: Pseudorandom number generator with individual assignments concealed in sequentially numbered sealed envelopes that were opened in order when assignments were made. Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Turkey Setting: Hospital (day‐case surgery) Sample size: 125 Number randomised: 62 to intervention (Modified Limberg flap) 63 to comparator (Modified Karydakis flap) Number analysed: 60 to intervention (Modified Limberg flap) 2 participants were lost 60 to comparator (Modified Karydakis flap) 3 participants were lost Baseline characteristics: Age (years, mean ± SD): Intervention group: 23.9 ± 5.4; median 23 (range 16 to 41) Control group: 22.6 ± 2.4; median 23 (range 15 to 36) Male (n, %): Intervention group: 58, 96.7% Control group: 54, 90% BMI (mean ± SD): Intervention group: 27.5 ± 3.5; median 27 (range 22 to 38) Control group: 28.5 ± 3.7; median 28 (range 23 to 40) Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 62 Control group: 63 Acute pilonidal abscesses Initially incised and drained under a local anaesthetic in the outpatient clinic, and antibiotics were prescribed. Inclusion criteria for the study: Patients presenting with chronic sacrococcygeal pilonidal sinus disease. Chronic disease was defined as “constant or recurrent symptoms observed over a period of several months without indication of spontaneous healing.” Exclusion criteria for the study: Patients who declined to undergo either procedure for cosmetic reasons were excluded from the study. Other exclusion criteria included patients with unilateral secondary openings located more than 2 cm from the midline, bilateral secondary openings, and secondary openings located near the anus.
Interventions	Intervention group: Off‐midline (modified Limberg flap) Comparator group: Off‐midline (modified Karydakis flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. A single preoperative intravenous injection of a third‐generation cephalosporin.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Operative time (minutes) (median, range) Other outcomes reported by the study: Other adverse effects Patient satisfaction Hypoesthesia Follow‐up: Patients were followed up by physical examination weekly for the first 4 postoperative weeks, at the end of months 3, 6, 9, and 12, and every 6 months thereafter. The median follow‐up duration was 20.5 (range 6 to 36) months.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Bukhari 2020.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: "The sample size consisted of 122 patients (each group having 61 cases) was estimated with confidence interval of 90 per cent, 10% absolute precision with a percentage of rhomboid excision and limberg flap repair as 85% and percentage of simple excision and primary closure as 89% for patients undergoing sacrococcygeal pilonidal sinus surgery." Randomisation: Probability simple random sampling technique Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Pakistan Setting: Hospital Sample size: 122 Number randomised: 61 to intervention group (Limberg Flap) 61 to control group (Conventional midline closure) Number analysed: 61 to intervention group (Limberg Flap) 61 to control group (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 28.23 ± 6.15 Control group: 27.49 ± 5.79 Male (n, %): Intervention group: 38, 62.3% Control group: 31, 50.8% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 61 Control group: 61 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients aged 20 to 40 years, either gender, having primary sacrococcygeal pilonidal sinus (presenting with acute symptoms). Exclusion criteria for the study: Patients having recurrent chronic pilonidal sinus disease (on medical record). Patients with bleeding disorder and type I or II diabetes.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Before the incision, a single shot of antibiotics was given in all cases.
Outcomes	Outcomes of interest in the review: Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) Other outcomes reported by the study: Other adverse effects Follow‐up: Intervention group: Not mentioned Control group: Not mentioned
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Can 2010.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: It was determined that 69 patients per treatment arm would allow for the detection of a 10% difference in the rate of early complications with 80% power and 5% confidence. Randomisation method: Block randomisation Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Turkey Setting: Hospital Sample size: 145 Number randomised: 72 to intervention group (Modified Limberg flap) 77 received intervention 73 to control group (Karydakis flap) 68 received intervention Number analysed: Primary outcome (early complication rate) 77 to intervention group (Modified Limberg flap) 68 to control group (Karydakis flap) Secondary (Recurrence) and tertiary (Patient satisfaction) outcome 74 to intervention group (Modified Limberg flap) Lost to follow‐up (N = 3) 63 to control group (Karydakis flap) Lost to follow‐up (N = 5) Baseline characteristics: Age (years, mean ± SD): Intervention group: 23.0 ± 4.4; median 22 (range 19 to 40) Control group: 22.5 ± 2.3; median 22 (range 20 to 31) Male (n, %): Intervention group: 76, 98.7% Control group: 67, 98.5% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 70 Control group: 63 Recurrence disease Intervention group: 7 Control group: 4 Unspecified Intervention group: 0 Control group: 1 Inclusion criteria for the study: Patients who presented with chronic symptomatic sacrococcygeal pilonidal sinus. Exclusion criteria for the study: Patients had sacrococcygeal pilonidal abscesses at the time of admission.
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Off‐midline (Karydakis flap) Use of additional interventions in both groups: Drains: Partially. Intervention group: 65 Control group: 23 Preoperative antibiotic: Yes. Prior to beginning the surgery, a single dose of cefazolin (1 g) was given intravenously. Methylene blue: Not mentioned.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Time from the operation to telephone contact (month) Patient satisfaction Follow‐up: mean 16.8 ± 3.4 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Daglar 2004.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned (the patients were randomly divided into two groups). Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 69 Number randomised: 37 to intervention (Limberg flap) 32 to comparator (Conventional midline closure) Number analysed: 37 to intervention (Limberg flap) 32 to comparator (Conventional midline closure) Baseline characteristics: The two groups were homogenised in terms of age, gender, occupation, body mass index (preoperative inflammation or presence of infection, preoperative antibiotic use, preoperative length of stay). Age (years, mean ± SD): Total: 25 (range 16 to 35) Male (n, %): Total: 58, 84% BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: 37 Control group: 32 Recurrence disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with chronic pilonidal sinus. Exclusion criteria for the study: Five patients in the primary repair group who refused this operation and requested a rhomboid flap were excluded from the study.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drains: Partially. Intervention group: 34 Control group: 8 While the drain usage rate was 25% (8 patients) in comparator group, it was 91.9% (34 patients) in intervention group (P < 0.05) Methylene blue: Not mentioned Perioperative antibiotic: Partially. All wound infections could be controlled with oral antibiotics.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (data not available) Recurrence rate Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) (data not available) Other outcomes reported by the study: Other adverse effects Mean drainage time Follow‐up: The mean follow‐up period was 28.4 (12 to 89) months in comparator group and 31.5 (13 to 52) months in intervention group.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Dass 2012.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: India Setting: Hospital Sample size: 80 Number randomised: 40 to intervention group (Limberg flap) 40 to control group (Conventional midline closure) Number analysed: 40 to intervention group (Limberg flap) 40 to control group (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD) Total: 28.4 ± 8.6 (range 16 to 48) Male (n, %): Intervention group: 38, 95% Control group: 36, 90% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 40 Control group: 40 Recurrence disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with sacrococcygeal pilonidal disease. Exclusion criteria for the study: Patients with recurrent disease and purulent discharge were excluded from the study.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drains: Not mentioned Methylene blue: Yes Perioperative antibiotic: Yes. All patients received ceftriaxone 1 g intravenously before shifting to the theatre.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Postoperative or post‐procedural pain (VAS, 1st postoperative day) (data not available) Other outcomes reported by the study: Other adverse effects Postoperative or post‐procedural pain (VAS, 2nd postoperative day) Follow‐up: Follow‐up was done after 1, 2, and 4 weeks and 3, 6, and 12 months for the first year after surgery. After the first year, patients were followed yearly for the subsequent 2 years to conclude a total of 3 years of follow‐up.
Funding sources	None
Declarations of interest	None
Notes	Publication status: Published

El Hadidi 2019.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: The randomisation results were put into opaque envelopes numbered sequentially and opened later by a healthcare professional not engaged in the operating room. Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Egypt Setting: Hospital Sample size: 120 Number randomised: 60 to intervention (Karydakis flap) 60 to comparator (Limberg flap) Number analysed: 60 to intervention (Karydakis flap) Analysis after follow‐up: 56 to intervention (Karydakis flap), lost to follow‐up (n = 4) 60 to comparator (Limberg flap) Analysis after follow‐up: 55 to comparator (Limberg flap), lost to follow‐up (n = 5) Baseline characteristics: Age (mean ± SD): Intervention group: 23.1 ± 4.8; median 22 (range 18 to 40); Control group: 22.8 ± 3.5; median 22 (range 19 to 35). Male (n, %): Intervention group: 57, 95% Control group: 58, 96.7% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 0 Control group: 0 Recurrent disease Intervention group: 60 Control group: 60 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: "Patients who were treated for recurrent SPD by either KF [Karydakis flap] or LF [Limberg flap] in our hospital from 2014 to 2017." Exclusion criteria for the study: Patients with a disease process that required unroofing and drainage procedures at the time of admission.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. A single dose of third‐generation cephalosporin was administered intravenously preoperatively. All patients were on postoperative antibiotic coverage for 5 days.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Follow‐up: Follow‐up examinations were made at the end of 1, 6, and 12 months after surgery. Beyond 1.5 years, the patients were monitored by phone calls at 6‐monthly intervals. Mean follow‐up period: 23 months, range 18 to 28 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Elhiny 2021.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Sample size was calculated according to Alvandipour 2019. Randomisation method: Block randomisation Blinding: The type of treatment was masked for the patient (single‐blinded technique). CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 54 Number randomised: Not mentioned Number analysed: 27 to intervention group (Karydakis flap) 27 to control group (Limberg flap) Number analysed: 27 to intervention group (Karydakis flap) 27 to control group (Limberg flap) Baseline characteristics: Age (years, mean ± SD): Intervention group: 33 ± 6.9 years (range 21.0 to 45.0) Control group: 29.8 ± 8.4 (range 19.0 to 48.0) Male (n, %): Intervention group: 12, 44.4% Control group: 11, 40.7% BMI (mean ± SD): Not mentioned Type of presentation Abscess Included, but no details Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: 27 Control group: 27 Inclusion criteria for the study: Patients with complicated or recurrent pilonidal sinus were included in the study. Exclusion criteria for the study: The exclusion criteria were cases with incomplete patients’ data, patients who were lost to follow‐up, patients presenting with different conditions mimicking pilonidal sinus, and patients with chronic conditions such as immunosuppression and diabetes mellitus.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drains: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Intravenous 1 g of ceftriaxone was administered as antibiotic prophylaxis.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate (at 12 months) Secondary outcomes: Adverse effects (wound infection, wound dehiscence (data not available)) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Analgesic use Cosmetic agreement Duration of drainage (weeks) Other adverse effects Painless toilet sitting (days) Patient satisfaction Time to first mobilisation Time to stitches removal (days) Time to normal daily activity (days) Follow‐up: Follow‐up was scheduled for all patients as visits at 3 and 10 days following discharge and once at the end of 1, 3, 6, and 12 months after the operation. The exclusion criteria includes patients who were lost to follow‐up. Consequently, the mean follow‐up duration for the cohort included in this study was estimated to be 12 months.
Funding sources	Not mentioned
Declarations of interest	None
Notes	Publication status: Published

Enshaei 2014.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 80 Number randomised: 40 to intervention group (Rotational Flap) 40 to control group (Conventional midline closure) Number analysed: 40 to intervention group (Rotational Flap) 40 to control group (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Total: 24.17± 6.05 (range 15 to 40) Male (n, %): Intervention group: 29, 72.5% Control group: 22, 55% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: 40 Control group: 40 Inclusion criteria for the study: Patients with chronic pilonidal sinus Exclusion criteria for the study: Diabetic and obese patients with underlying diseases and patients with acute pilonidal abscess or prior surgery were excluded.
Interventions	Intervention group: Off‐midline (Rotational flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Not mentioned Methylene blue: Not mentioned Perioperative antibiotic: Yes. Antibiotics were administered to patients before and after surgery.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (time to suture removal) (days) Recurrence rate (6 months) Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Duration of postoperative pain Follow‐up: "Patients discharged and the first and fourth weeks and 6 months postoperatively were visited and questionnaires were completed."
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Enshaei 2020.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A number of sequences were assigned by an independent researcher using random allocation software version 1 in permuted blocks size of 3 and 6. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 100 Number randomised: 50 to intervention group (Rotational flap) 50 to control group (Conventional midline closure) Number analysed: 47 to intervention group (Rotational flap) Lost to follow‐up (N = 3) 49 to control group (Conventional midline closure) Lost to follow‐up (N = 1) Baseline characteristics: Age (years, mean ± SD) Intervention group: 27.32 ± 4.5 Control group: 25.84 ± 5.5 Male (n, %): Intervention group: 41, 87.2% Control group: 43, 87.8% BMI (mean ± SD): Intervention group: 26.8 ± 1.9 Control group: 27.3 ± 2.2 Type of presentation Primary disease Intervention group: 47 Control group: 49 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: "Patients with non‐recurrent pilonidal sinus without any sign of infection giving informed consent to participate in the study." Exclusion criteria for the study: "Patients with acute pilonidal sinus along with abscess and chronic recurrent pilonidal sinus, as well as patients with diabetes, those receiving corticosteroids and obese subjects. Patients with BMI < 30 kg/m2 were defined as obese patients and excluded from the study."
Interventions	Intervention group: Off‐midline (Rotational flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drains: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Prior to operation, a single dose of antibiotic (cefazolin 1 g) was injected intravenously.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) Other outcomes reported by the study: Other adverse effects Mean duration of pain relief after surgery Follow‐up: "Following the discharge, we asked the patients to visit the outpatient clinic at the end of the first, second and fourth weeks, as well as 3 and 6 months after surgery." "four patients did not refer for follow‐up and were excluded." Consequently, the mean follow‐up duration for the cohort included in this study was estimated to be 6 months.
Funding sources	Not mentioned
Declarations of interest	The authors declare that they have no conflicts of interest.
Notes	Publication status: Published

Ertan 2005.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Allocation was performed by balanced randomisation after informed consent had been obtained. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 100 Number randomised: 50 to intervention (Limberg flap) 50 to comparator (Conventional midline closure) Number analysed: 50 to intervention (Limberg flap) 50 to comparator (Conventional midline closure) Baseline characteristics: Age (mean ± SD): Intervention group: 28.0 ± 7.7 Control group: 26.3 ± 6.8 Male (n, %): Total: 91 (91%) male, 9 (9%) female Intervention group: 46, 92% Control group: 45, 90% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 50 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with surgery for sacrococcygeal pilonidal sinus. Exclusion criteria for the study: Participants with acute abscesses or recurrent sinuses.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) (data not available) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (data not available) Length of hospital stay (days) (data not available) Quality of life (SF‐36, at 12 months) Other outcomes reported by the study: Postoperative or post‐procedural pain (VAS) (time point was not reported) Other adverse effects Time to sitting on toilet without pain Time to walk without pain Follow‐up: "By the time of data analysis (February 2004), 73 patients were at 1 year after surgery, and we were able to interview 65 of them (31 in group I and 34 in group II) by telephone about their satisfaction with treatment and any recurrences. Total follow period: mean 19 months (range 4 ‐ 38)."
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Galal 2012.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Randomisation was achieved through a computer‐generated schedule and the results were sealed into envelopes. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 120 Number randomised: 60 to intervention (Modified Limberg flap) 60 to comparator (Conventional midline closure) Number analysed: 60 to intervention (Modified Limberg flap) 60 to comparator (Conventional midline closure) Baseline characteristics: Age (mean years ± SD): Intervention group: 26 ± 8.6 Control group: 27 ± 9.4 Male (n, %): Intervention group: 45, 75% Control group: 46, 76.7% BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: 60 Control group: 60 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with uncomplicated pilonidal disease. Exclusion criteria for the study: "The patients who presented with acute pilonidal abscesses were excluded from this study. Recurrent or complex pilonidal sinuses were also excluded from this study."
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: No Perioperative antibiotic: Yes. Antibiotic prophylaxis of 1.2 g amoxicillin–clavulinic acid was given intravenously at the time of induction and continued for 48 h. Then this was changed to oral form for five days.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Postoperative or post‐procedural pain on the first postoperative day (VAS, 1st postoperative day) Other outcomes reported by the study: Other adverse effects Follow‐up: Follow‐up (months) of intervention group: 22 ± 7.64 Follow‐up (months) of control group: 21.5 ± 6.82
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Guner 2013.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: A computer‐assisted random number table Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Turkey Setting: Hospital Sample size: 122 Number randomised: 61 to intervention (Bascom’s cleft lift) 61 to comparator (Limberg flap) Number analysed: 61 to intervention (Bascom’s cleft lift) 61 to comparator (Limberg flap) Baseline characteristics: Age (mean years): Intervention group: 24.64 (range 18 to 46) Control group: 25.41 (range 18 to 48) Male (n, %): Intervention group: 50, 82.0% Control group: 48, 78.7% BMI (mean): Intervention group: 24.93 Control group: 25.76 Type of presentation Primary disease Intervention group: not mentined Control group: not mentined Recurrent disease Intervention group: not mentined Control group: not mentined Unspecified Intervention group: 61 Control group: 61 Inclusion criteria for the study: The main inclusion criterion was the presence of pilonidal sinus disease in a patient ≥ 18 years of age. Exclusion criteria for the study: Patients presenting with acute abscess formation (dirty wounds according to the modified Cruse‐Foord classification) or active psychiatric disease were excluded as were patients who refused to take part in the study.
Interventions	Intervention group: Off‐midline (Bascom’s cleft lift) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Before the operation, a single‐dose of cefazolin 1 g was administered intravenously as a prophylactic antibiotic.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) (Data are expressed as the mean and range) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Quality of life (SF‐36, on days 10 and 30) Operative time (minutes) (Data are expressed as the mean and range) Length of hospital stay (days) (Data are expressed as the mean and range) Other outcomes reported by the study: Other adverse effects Postoperative or post‐procedural pain (VAS, on postoperative days 2, 10, 30) Weight of removed tissue Follow‐up: The patients came for follow‐up visits on postoperative days 2, 10, and 30 and then once every 6 months. Median period: 13 months. Intervention group: 13 (range 6 to 18) Control group: 13 (range 6 to 18)
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Hemmingsson 2022.

Study characteristics
Methods	Study design: Parallel‐group RCT (early termination of the study based on the results of a planned interim analysis of the first 100 participants recruited) Sample size calculation: Not mentioned Randomisation method: The randomisation sequence was computer‐generated using a random number technique and concealed. Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Sweden Setting: Hospital (day‐case surgery) Sample size: 125 Number randomised: 64 to intervention (Karydakis flap) 4 did not receive allocated intervention 61 to comparator (Tension‐free midline closure) 5 did not receive allocated intervention Number analysed: 60 to intervention (Karydakis flap) 56 to comparator (Tension‐free midline closure) Baseline characteristics: Age (mean years ± SD): Intervention group: median 26.5 (95% CI 23 to 30) Control group: median 28 (95% CI 23 to 29) Male (n, %): Intervention group: 52, 86.7% Control group: 51, 91.1% BMI (mean ± SD): Intervention group: 27.1 ± 0.6 (range 26 to 29) Control group: 26.0 ± 0.4 (range 25 to 27) Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 60 Control group: 56 Inclusion criteria for the study: Adult patients aged 18 years of age or older with primary or once‐recurrent pilonidal disease were considered for surgical treatment. Patients were included if at least 4 weeks had elapsed since any acute incision for a pilonidal abscess. Exclusion criteria for the study: Patients considered for only minimally invasive procedures (shaving, picking loose hair from skin openings, or minimal curettage) were excluded.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: No. An infected wound was sufficiently opened to drain any infected wound cavities, followed by washing out with saline. Methylene blue or dye: No. Perioperative antibiotic: Yes. All patients received antibiotic prophylaxis with a single dose of 1 g metronidazole and 160 mg sulfamethoxazole combined with 800 mg trimethoprim as an intravenous injection prior to surgery.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) (data not available for meta‐analysis, median, 95% CI) Proportion of wounds healed (3 months, 12 months) Recurrence rate Lost to follow‐up before 30 days (Intervention group: N = 3; control group: N = 5). Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) (data not available, median, 95% CI) Operative time (minutes) (data not available, median, 95% CI) Other outcomes reported by the study: Other adverse effects Follow‐up: Median follow‐up time was 3940.5 (range 11 to 5069) days. Participants were followed up until complete wound healing; late follow‐up after 6 to 13 years was performed by telephone by two blinded assessors.
Funding sources	Research grants by County Council of Västerbotten, Umeå, Sweden, and County Council of Norrbotten, Luleå, Sweden. The funding sources were not involved in the conduct of the study or report.
Declarations of interest	Not mentioned
Notes	Publication status: Published

Kartal 2019.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: A system of sequentially numbered sealed envelopes. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital (day‐case surgery) Sample size: 100 Number randomised: 50 to intervention (Modified Karydakis flap) 50 to comparator (Karydakis flap) Number analysed: 50 to intervention (Modified Karydakis flap) 50 to comparator (Karydakis flap) Baseline characteristics: Age (mean years ± SD): Intervention group: 24.28 ± 7.40 Control group: 24.78 ± 7.00 Male (n, %): Intervention group: 42, 84% Control group: 41, 82% BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: 50 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with planned pilonidal sinus surgery. The number of sinuses in the sacrococcygeal region and their distances to each other were not considered. Exclusion criteria for the study: Patients who did not accept the procedure (n: 9), had undergone previous pilonidal sinus surgery (n: 7), had active infection (n: 9), and were not minimum of 18 years of age (n: 4) were excluded.
Interventions	Intervention group: Off‐midline (Modified Karydakis flap) Comparator group: Off‐midline (Karydakis flap) Use of additional interventions in both groups: Drain: No Methylene blue: Not mentioned Perioperative antibiotic: Yes. During the induction of anaesthesia, 1 gram of cefazolin sodium was administered intravenously.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Follow‐up: The mean follow‐up period: 25 (range 18 to 33) months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Khan 2013.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A table of random numbers. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: India Setting: Hospital Sample size: 120 Number randomised: 60 to intervention (Limberg flap) 60 to comparator (Conventional midline closure) Number analysed: 60 to intervention (Limberg flap) 60 to comparator (Conventional midline closure) Baseline characteristics: Age (mean years ± SD): Intervention group: 24 (range 17 to 42) Control group: 26 (range 16 to 40) Male (n, %): Intervention group: 53, 88.3% Control group: 51, 85.0% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 60 Control group: 60 Inclusion criteria for the study: Patients with pilonidal sinus in the sacrococcygeal area. Exclusion criteria for the study: Not mentioned.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Yes Perioperative antibiotic: Yes. Patients were given IV antibiotics for first 24 h followed by oral antibiotics for next 9 days.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (data not available) Length of hospital stay (days) (data not available) Operative time (minutes) (data not available) Other outcomes reported by the study: Not mentioned Follow‐up: The patients were regularly followed up for a period of 2 years.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Kohla 2016.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: No details Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 30 Number randomised: 15 to intervention (Karydakis flap) 15 to comparator (Limberg flap) Number analysed: 15 to intervention (Karydakis flap) 15 to comparator (Limberg flap) Baseline characteristics: Age (years, range): Intervention group: 18 to 25 Control group: 19 to 32 Male (n, %): Not mentioned BMI (mean ± SD): Intervention group: obese (n = 6), overweight (n = 4); Control group: obese (n = 5), overweight (n = 4). Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 15 Control group: 15 Inclusion criteria for the study: Patients with sacrococcygeal pilonidal sinus disease Exclusion criteria for the study: Not mentioned
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Yes Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Proportion of wounds healed Recurrence rate Secondary outcomes: Adverse effects (wound dehiscence) Time to return to work (days) Operative time (minutes) Other outcomes reported by the study: Time to removal of stitches Follow‐up: Not mentioned
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Maghsudi 2020.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A block randomisation method (including quadruple blocks) was used for this purpose. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 100 Number randomised: 50 to intervention (Limberg flap) 50 to comparator (Conventional midline closure) Number analysed: 50 to intervention (Limberg flap) 50 to comparator (Conventional midline closure) Baseline characteristics: Age (mean years ± SD): Intervention group: 24.0 ± 4.6 (range 14 to 36) Control group: 24.1 ± 4.5 (range 14 to 34) Male (n, %): Intervention group: 29, 58% male Control group: 33, 66% male BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 50 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with primary pilonidal cysts or abscesses. Exclusion criteria for the study: Patients with immune deficiency disease or diabetes were excluded from the study.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. All patients received one dose of prophylactic antibiotic (i.e. 1 g of intravenous cefazolin); patients were discharged with a prescription for 500 mg of oral cephalexin to be taken every 6 hours.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) (median ± SD) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (median ± SD) Operative time (minutes) (median ± SD) Other outcomes reported by the study: Ability to sit on toilet Mobilisation Postoperative or post‐procedural pain (VAS) (time point was not reported) Other adverse effects Follow‐up: All patients returned to the clinic after 1 week, 1 month, and 3 months for follow‐up.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Muzi 2010.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Eligible patients were assigned randomly to 1 of the 2 surgical groups according to a computer‐generated list. Allocations were sealed in opaque numbered envelopes. Blinding: An independent observer, who was not from the surgical team and who was unaware of the treatment assignments, recorded all data, which included postoperative events and follow‐up findings. CONSORT flow diagram: Yes
Participants	Location: Italy Setting: Hospital (midline closures were day‐case procedures) Sample size: 272 Number randomised: 135 to intervention (Limberg flap) 137 to comparator (Conventional midline closure) Number analysed: 130 to intervention (Limberg flap) Lost at follow‐up evaluation (n = 5) 130 to comparator (Conventional midline closure) Lost at follow‐up evaluation (n = 7) Baseline characteristics: Age (mean years ± SD): Intervention group: 25.05 ± 7.53 Control group: 25.01 ± 8.58 Male (n, %): Intervention group: 114, 88% Control group: 109, 84% BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 130 Control group: 130 Inclusion criteria for the study: Patients with pilonidal sinus disease, older than age 14. Exclusion criteria for the study: ASA group higher than III; risk of malignant hyperthermia; obesity (body mass index, > 35); insulin‐dependent diabetes; epilepsy (not well controlled); severe allergy toward local anesthetics or other medications; social‐familiar reasons such as a lack of assistance at home after surgery for day surgery patients; abuse of substances with central nervous system activity; drug addiction; and alcoholism.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Intervention group: Yes Control group: No Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence (partial and complete)) Time to return to work (days) Length of hospital stay (days) Other outcomes reported by the study: Success of treatment Days of confinement in bed Postoperative or post‐procedural pain (VAS, 3rd postoperative day) Follow‐up: The patients were followed up with clinical controls until the complete healing of the wound. Subsequently, the patients were followed up with telephone interviews. The mean follow‐up period of intervention group: 45.74 ± 7.6 months The mean follow‐up period of control group: 47.83 ± 6.6 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published The authors of Muzi 2010 referred to their method as "Muzi’s tension‐free primary closure" (Muzi 2018). However, the technique they describe ("In group A, patients were placed in the prone position; the area to be excised was marked on the skin, and a rhomboid incision to the presacral fascia, including the sinus and its extensions, was made down. The cavity was then closed with absorbable sutures in a double layer with alternating deep and superficial sutures to minimize tissue traction and skin edge impairment. Thus, the skin edges were approximated without tension using interrupted sutures of absorbable 3/0 suture or liquid skin adhesive.") does not match or closely resemble the description of "tension‐free midline closure" as described in the German National Guideline on the management of pilonidal disease: update 2020 (Iesalnieks 2021), which states, "The technique includes a subcutaneous release at the level of the fascia of the gluteus maximus muscle as wide as 3–5 cm, leading to the flattening of the wound after midline closure." This method should be a type of modified midline closure method. Therefore, we categorised this study into the conventional midline closure group.

Okuş 2012.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A closed envelope method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 100 Number randomised: Not mentioned. Seven participants were lost during follow‐up and were excluded from the study. Number analysed: 49 to intervention group (Limberg flap) 44 to control group (Tension‐free midline closure) Baseline characteristics: Age (mean years ± SD): Total: 25.6 ± 5.3; median 25 (range 17 to 43) Intervention(Limberg flap): 24.2 ± 3.6; median 24 (range 17 to 33) Comparator (tension‐free primary closure): 26.3 ± 5.9; median 25.5 (range 17 to 43) Male (n, %): Intervention (Limberg flap): 48, 98.0% Comparator (Tension‐free midline closure): 40, 90.9% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: 49 Control group: 44 Inclusion criteria for the study: Patients with pilonidal disease between May 2006 and August 2010. Exclusion criteria for the study: Patients who expressed preference for one technique were not included in the study.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) (data not available) Other outcomes reported by the study: Not mentioned Follow‐up: The patients returned for follow‐up visits at 10 days and at 1, 3, and 6 months; they were then seen yearly. The median follow‐up period was 29.5 months (8 to 43 months). Limberg flap: 31.5 (8 to 42) months Tension‐free primary: 28 (8 to 43) months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Rashidian 2014.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: The block randomisation method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 40 Number randomised: 20 to intervention (Limberg flap) 20 to comparator (Tension‐free midline closure) Number analysed: 20 to intervention (Limberg flap) 20 to comparator (Tension‐free midline closure) Baseline characteristics: Age (mean ± SD ): Not mentioned Male (n, %): Not mentioned BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: 20 Control group: 20 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients referred to the surgical clinics of Firoozgar general hospital with a diagnosis of primary sacrococcygeal pilonidal disease were included. Exclusion criteria for the study: Patients with acute infection, abscess, or recurrent pilonidal disease at the time of referral were excluded.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. All the patients received 1 g intravenous cefazolin as prophylaxis one hour before the operation.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) (data not available) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (data not available) Length of hospital stay (days) (data not available) Other outcomes reported by the study: Other adverse effects Follow‐up: The mean duration for follow‐up was 18 months (range: 3 months to 3 years). Follow‐up was performed at 1 week, 2 weeks, 1 month and 6 months after the surgery.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Riaz 2019.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Patients were randomly divided into two equal groups of n patients each by lottery method. Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Pakistan Setting: Hospital Sample size: 204 Number randomised: 102 to intervention (Karydakis flap) 102 to comparator (Limberg flap) Number analysed: 102 to intervention (Karydakis flap) 102 to comparator (Limberg flap) Baseline characteristics: Age (mean years ± SD ): Total: 27.21 ± 8.24 Intervention group: 26.36 ± 7.69 Control group: 28.06 ± 8.71 Male (n, %): Intervention group: 80, 78.4% Control group: 86, 84.3% BMI (mean ± SD ): Not mentioned Type of presentation Primary disease Intervention group: 102 Control group: 102 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients of either gender presenting with saccrococcygeal pilonidal sinus; with age ranging from 15 to 60 years, and having ASA grade I to III. Exclusion criteria for the study: All those patients who presented with a concurrent abscess; recurrent pilonidal sinus; concurrent perianal pathology like fistula in ano; ASA grade IV and V; diabetes mellitus and compromised immune status were excluded from the study sample.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Off‐midline (Karydakis flap) Use of additional interventions in both groups: Drain: Yes. Both groups of patients had a Redivac suction drain placed. Methylene blue: Yes. Methylene blue dye was injected to delineate all the tracks of the pilonidal sinus. Perioperative antibiotic: Yes. A dose of 1.2 grams of Augmentin (amoxicillin and clavilunate) as a prophylactic antibiotic was administered before anaesthesia.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Postoperative or post‐procedural pain (VAS, 28th postoperative day) Follow‐up: Mean duration of 6 months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Sakr 2006.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: The closed envelope method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 170 Number randomised: Total: 170 Lost to follow up (n = 9) Number analysed: 79 to intervention (Karydakis flap) 82 to comparator (Conventional midline closure) Baseline characteristics: Age (mean ± SD, years): Intervention group: 26.5 (range 18 to 43) Control group: 22.7 (range 14 to 37) Male (n, %): Intervention group: 68, 86.1% Control group: 70, 85.4% BMI (mean ± SD): Intervention group: 29.1 Control group: 27.4 Type of presentation Primary disease Intervention group: 79 Control group: 82 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with chronic pilonidal sinus admitted to the department of surgery of Ahmadi Hospital between June 2001 and June 2004. Exclusion criteria for the study: Participants with acute abscesses or recurrent sinuses.
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes Methylene blue: Yes Perioperative antibiotic: Yes. One dose of intravenous second‐generation cephalosporin was given to all patients with induction of anaesthesia.
Outcomes	Outcomes of interest in the review: Primary outcomes: Proportion of wounds healed (11th and 21st postoperative day) Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Time to return to work (days) (data not available) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Follow‐up: Patients were followed up for a period that ranged from 12 months to 49 months with a mean of 30.4 months.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Saydam 2015.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: The patients were randomly divided into two groups by the computer. Blinding: Not mentioned CONSORT flow diagram: No
Participants	Location: Turkey Setting: Hospital Sample size: 100 Number randomised: 50 to intervention (Modified Limberg flap) 50 to comparator (Burow’s triangle advancement flap) Number analysed: 50 to intervention (Modified Limberg) 50 to comparator (Burow’s triangle advancement flap) Baseline characteristics: Age (years, mean ± SD): Intervention group: median 23 (range 19 to 34) Control group: median 23 (range 20 to 34) Male (n, %): Intervention group: 50, 100% Control group: 50, 100% BMI (mean ± SD): Intervention group: median (range) 24.7 (21.0 to 29.0) Control group: median (range) 25.2 (20.0 to 36.0) Type of presentation Primary disease Intervention group: 50 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with PSD admitted to Ankara Mevki Military Hospital General Surgery Service between May 2013 and August 2013. Exclusion criteria for the study: Age below 15 years, collagen tissue disease, recurrent generalized PSD, and infection or abscess in relation to PSD.
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Off‐midline (Burow’s triangle advancement flap) Use of additional interventions in both groups: Drain: Yes Methylene blue: Not mentioned Perioperative antibiotic: Yes. Administration of a single intravenous dose of a broad‐spectrum antibiotic as prophylaxis was performed on the day of surgery.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate (12 months) Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) (median, range) Operative time (minutes) (median, range) Other outcomes reported by the study: Postoperative or post‐procedural pain (time point was not reported) Follow‐up: The patients were asked to come for follow‐up appointments on the 10th postoperative day, especially for wound infection, and the first, sixth, and twelfth month postoperatively. The mean follow‐up period for all cases was 12 months.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published The study asserted that it was registered on ClinicalTrials.gov under the identifier NCT02116738. However, upon examination, the information provided on ClinicalTrials.gov does not correspond with the details of this study.

Sevinc 2016.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: A system of sequentially numbered sealed envelopes. Blinding: Not mentioned CONSORT flow diagram: Yes
Participants	Location: Turkey Setting: Hospital Sample size: 150 Number randomised: 50 to intervention 1 (Limberg flap) 50 to intervention 2 (Karydakis flap) 50 to comparator (Tension‐free midline closure) Number analysed: 50 to intervention 1 (Limberg flap) 50 to intervention 2 (Karydakis flap) 50 to comparator (Tension‐free midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention 1 group (Limberg flap): 25.0 ± 6.2; median 23.5 (range 18 to 46) Intervention 2 group (Karydakis technique): 24.7 ± 4.5; median 24 (range 18 to 38) Control group: 25.0 ± 6.5; median, 23.5 (range 18 to 47) Male (n, %): Intervention 1 group (Limberg flap): 39, 78% Intervention 2 group (Karydakis technique): 41, 82% Control group: 43, 86%. P = 0.582 BMI (mean ± SD): Intervention 1 group (Limberg flap): 24.1 ± 3.26 Intervention 2 group (Karydakis technique): 24.5 ± 2.81 Control group: 24.3 ± 2.62. P = 0.337 Type of presentation Primary disease Intervention group: 100 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients with primary pilonidal sinus. Exclusion criteria for the study: Participants with recurrent sinuses.
Interventions	Intervention 1 group Off‐midline (Limberg flap) Intervention 2 group Off‐midline (Karydakis flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: yes. Methylene blue: Not mentioned Perioperative antibiotic: No. Prophylactic antibiotics were not used by any of the patients.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Operative time (minutes) (median, range) Other outcomes reported by the study: Other adverse effects Follow‐up: The patients were followed up after the first, sixth, and twelfth month in the outpatient clinic. After the first year after surgery, the patients were called by phone annually. The median (range) follow‐up time: 24.2 (18.5 to 34.27) months
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Shabbir 2014.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: The lottery method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Pakistan Setting: Hospital Sample size: 60 Number randomised: 30 to intervention (Modified Limberg flap) 30 to comparator (Conventional midline closure) Number analysed: 30 to intervention (Modified Limberg flap) 30 to comparator (Conventional midline closure) Baseline characteristics: Age: Not mentioned Male (n, %): Intervention group: 27, 90% Control group: 28, 93.3% BMI: Not mentioned Type of presentation Primary disease Intervention group: 30 Control group: 30 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients over 13 years of age with sacrococcygeal pilonidal sinus were included. Exclusion criteria for the study: Patients with recurrent pilonidal sinus.
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Yes. Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate (12 months) Secondary outcomes: Adverse effects (wound infection) Time to return to work (days) Length of hospital stay (days) Other outcomes reported by the study: Other adverse effects Follow‐up: Patients were followed up for 12 months for surgical complications of the treatment.
Funding sources	Not mentioned
Declarations of interest	Not mentioned
Notes	Publication status: Published

Tavassoli 2011.

Study characteristics
Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: No details Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 100 Number randomised: 50 to intervention (Limberg flap) 50 to comparator (Tension‐free midline closure) Number analysed: 50 to intervention (Limberg flap) 50 to comparator (Tension‐free midline closure) Baseline characteristics: Age (years, mean ± SD): Total: 24 ± 4.35 Male (n, %): Total: 80, 80% BMI: Not mentioned Type of presentation Primary disease Intervention group: 50 Control group: 50 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Inclusion criteria: people between 15 and 45 years old; written consent of the patients for surgical treatment after being informed about the operative procedure and type. Exclusion criteria for the study: 1. Abscess formation; 2. Immunodeficiency and diabetes mellitus; 3. Patients younger than 15 and older than 45 years old; 4. Existing recurrent disease or previous surgery in the sacrococcygeal region; 5. Severe hirsutism in female patients; 6. Patients with psychiatric disease or poor hygiene.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Tension‐free midline closure Use of additional interventions in both groups: Drain: Off‐midline (Limberg flap): Yes Tension‐free primary closure: Not mentioned Methylene blue: Not mentioned Perioperative antibiotic: Yes. They received 1 g cefazolin 30 minutes before the incision.
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (time to suture removal) (days) (data not available for meta‐analysis) Recurrence rate (6 months) Secondary outcomes: Adverse effects (wound infection) Time to return to work (days) (data not available for meta‐analysis) Operative time (minutes) (data not available for meta‐analysis) Postoperative or post‐procedural pain (VAS, 1st postoperative day) Other outcomes reported by the study: Other adverse effects Postoperative or post‐procedural pain (VAS, 4th postoperative day) Follow‐up: 6 months.
Funding sources	Faculty of Medicine of Mashhad University of Medical Sciences
Declarations of interest	Not mentioned
Notes	Publication status: Published

ASA: American Society of Anaesthesiologists; BMI: body mass index; IQR: interquartile range; PSD: pilonidal sinus disease; RCT: randomised controlled trial; SD: standard deviation; SF‐36: 36‐item short form quality of life questionnaire; SPD: sacrococcygeal pilonidal disease; VAS: visual analogue scale

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Ates 2011	Quasi‐randomised trial: participants were assigned to groups using patient admission numbers as the basis for randomisation.
Awad 2007	Ineligible study design: the study design appears to be flawed, as the authors only assert that it is a randomised controlled trial within the abstract.
Cihan 2006	Quasi‐randomised trial: the randomisation of the patients into groups was done in such a way that every other patient with pilonidal disease was assigned to the group with a modified method of treatment.
Ersoy 2009	Quasi‐randomised trial: participants were assigned to groups using patient admission numbers as the basis for randomisation.
Karim 2020	Quasi‐randomised trial: this study stated that it was based on an interventional quasi‐experimental design.
Kayal 2014	Quasi‐randomised trial: participants were assigned to groups using hospital registration numbers as the basis for randomisation.
Nordon 2009	Ineligible comparison. In this publication, the "Simple Bascom technique" refers to a pit‐picking procedure in which the abscess cavity is drained through a generous lateral incision, subsequently left open to heal by secondary intention (Smart 2013). This technique is typically regarded as a minimally invasive surgical approach (Iesalnieks 2019), and should not be considered as a midline or off‐midline wound closure method. This specific comparison is unique to this study, and another Cochrane Review has already assessed this procedure (Al‐Khamis 2010).
Tokac 2015	Quasi‐randomised trial: participants were assigned to groups using patient admission numbers as the basis for randomisation.
Zorlu 2016	Quasi‐randomised trial: participants were assigned to groups using outpatient attendance numbers as the basis for randomisation.

Characteristics of studies awaiting classification [ordered by study ID]

Ahmad 2018.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: India Setting: Hospital Sample size: 100 Number randomised: 50 to intervention (Modified Limberg flap) 50 to comparator (Conventional midline closure) Number analysed: 50 to intervention (Modified Limberg flap) 50 to comparator (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 41.6 (range 38 to 48); Control group: 34.8 (range 15 to 50). Male (n, %): Intervention group: 43, 86.0%; Control group: 45, 90.0%. BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 50 Control group: 50 Inclusion criteria for the study: Participants with pilonidal disease Exclusion criteria for the study: Participants with acute abscesses
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drains: both yes Methylene blue: both yes Perioperative antibiotic: both yes. Antibiotic prophylaxis of 1.2 g amoxicillin‐clavulinic acid was given intravenously at the time of induction and continued for 48 hours. Then this was changed to oral antibiotic form for five days.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence (2 years) Secondary outcomes: Adverse effects (wound infection) Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Other adverse effects Follow‐up: All patients were advised to visit the outpatient clinic after 1 week, then after 15 days, monthly for 3 months and 3 monthly for 3 years. Maximum follow‐up was up to 36 months.
Notes	Publication status: Published

Bayhan 2020.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Turkey Setting: Hospital Sample size: 104 Number randomised: Not mentioned Number analysed: Not mentioned Type of presentation: Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: not mentioned Control group: not mentioned Baseline characteristics: Age (years, mean ± SD): Not mentioned Male (n, %): Not mentioned BMI (mean ± SD): not mentioned Inclusion criteria for the study: Not mentioned Exclusion criteria for the study: Not mentioned
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Off‐midline (Bascom cleft lift) Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Wound healing Recurrence Secondary outcomes: Adverse effects Time to return to work (days) Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Time to removal of the vacuum drain Follow‐up: Patients were followed up in the postoperative 1st week, 1st month, 6th month, and 1st year.
Notes	Not mentioned

IRCT20180719040525N1.

Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 75 Calculation method: Not reported Randomisation: Randomised Randomisation description: "Randomization process is BBR (Balanced Block Randomization), in according to that patient are devided to 3 groupe with 25 members, then making 6 Block with alphabet A, B, C like (ABC,ACB,CBA,CAB,BCA,BAC). Then blocks had been lined for 25 times in according to Randomized numeral schedule." Blinding: Not blinded Other design features: Not reported
Participants	Location: Iran Setting: Rajaee hospital Velayat hospital Inclusion criteria for the study: Aged between 12 and 45 Uncomplicated sacrococcygeal pilonidal sinus Individual consents to participate in the study Exclusion criteria: Infection or abscess in pilonidal sinus History of pilonidal sinus surgery Individual does not give consent to participate Baseline characteristics: Age: 12 to 45 years old Gender: Both BMI: Not reported
Interventions	Intervention group 1: Off‐midline (Limberg flap): 25 Intervention group 2: Marsupialisation surgery: 25 Intervention group 3: Muzi tension‐free primary closure: 25 Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Wound healing time Time point: within 1 month after surgery Method of measurement: Clinical examination Recurrence Time points: 1 and 3 months after surgery Method of measurement: Physical examination Secondary outcomes: Degree of pain Time points: 12 and 24 hours after surgery Method of measurement: VAS measurement Other outcomes reported by the study: Not reported
Notes	Funding source Name: Vice Chancellor for Research at Qazvin University of Medical Sciences Type of organisation providing the funding: Academic Full name of funding responsible person: Dr Amir Peymani Grant name/code: Not reported

Kaleem 2021.

Methods	Study design: Parallel‐group RCT Sample size calculation: "367 sample size was calculated with 95% confidence interval, 80% power of study and taking expecting anticipated population proportion 3.97%" Randomisation: A lottery method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Pakistan Setting: Hospital Sample size: 90 Number randomised: 45 to intervention (Limberg flap) 45 to comparator (Conventional midline closure) Number analysed: 45 to intervention (Limberg flap) 45 to comparator (Conventional midline closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 44.37 ± 15.42 (range 20 to 70) Control group: 45.24 ± 14.50 (range 20 to 68) Male (n, %): Intervention group: 23, 51.1% Control group: 25, 55.6% BMI: Intervention group: normal BMI 19 (42.2%), overweight 12 (26.7%), obese 14 (31.1%) Control group: normal BMI 14(31.1%), overweight 20 (44.4%), obese 11 (24.4%) Type of presentation Primary disease Intervention group: not mentioned Control group: not mentioned Recurrent disease Intervention group: not mentioned Control group: not mentioned Unspecified Intervention group: 45 Control group: 45 Inclusion criteria for the study: Patients aged 20 to 70 years undergoing elective surgery for sacrococcygeal pilonidal sinus under spinal anaesthesia were included. Exclusion criteria for the study: Patients with diabetes mellitus, hypertension, renal disease, liver disease, hepatitis B and C, and cirrhosis were excluded.
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Conventional midline closure Use of additional interventions in both groups: Drain: Not mentioned Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Not mentioned Secondary outcomes: Adverse effects (wound infection) Length of hospital stay (days) Other outcomes reported by the study: Not mentioned
Notes	Publication status: Published

Khalil 2019.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 40 Number randomised: 20 to intervention group (off‐midline (modified Limberg flap)) 20 to control group (off‐midline (Bascom cleft lift)) Number analysed: Not mentioned Type of presentation: Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: Not mentioned Control group: Not mentioned Baseline characteristics: Age (years, mean ± SD): Not mentioned Male (n, %): Not mentioned BMI (mean ± SD): not mentioned Inclusion criteria for the study: This prospective study enroled 40 patients with complex or recurrent SPD presenting to Alexandria University Hospital. Exclusion criteria for the study: Not mentioned
Interventions	Intervention group: Off‐midline (modified Limberg flap) Comparator group: Off‐midline (Bascom cleft lift) Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Wound healing rate Secondary outcomes: Adverse effects Operative time (minutes) Postoperative pain Other outcomes reported by the study: Patient satisfaction score Functional recovery time
Notes	Poster presentation without data

Khan 2016.

Methods	Study design: Parallel‐group RCT Sample size calculation: "Sample size of 180 cases (90 in each group) is calculated with 80% power of test and 95% confidence level while taking expected percentage of seroma formation to be 4.2% with modified Limberg flap and 17.1% with modified Karydakis flap procedure in patients with sacrococcygeal pilonidal sinus." Randomisation method: A lottery method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Pakistan Setting: Hospital Sample size: 180 Number randomised: 90 to intervention (Modified Karydakis flap) 90 to comparator (Modified Limberg flap) Number analysed: 90 to intervention (Modified Karydakis flap) 90 to comparator (Modified Limberg flap) Baseline characteristics: Age (years, mean ± SD): Intervention group: 43.42 ± 16.66 Control group: 42.71 ± 14.78 Male (n, %): Total: 117, 65% Intervention group: 64, 71.1% Control group: 53, 58.9% BMI (mean ± SD): Not mentioned Type of presentation Primary disease Intervention group: 90 Control group: 90 Recurrent disease Intervention group: 0 Control group: 0 Unspecified Intervention group: 0 Control group: 0 Inclusion criteria for the study: Patients of both genders, aged between 18 and 70 years, with sacrococcygeal pilonidal sinus disease (as per operational definition) for ≤ 1 year. Patients who gave written informed consent. Exclusion criteria for the study: Patients with diabetes, anaemia, or on steroid therapy (history and clinical record). Patients with other concomitant disorders such as peri‐anal fissure (breach in the continuity of peri‐anal skin)/sinus (discharging opening in the peri‐anal area)/abscess (red, hot, and tender collection in the peri‐anal area) on clinical examination. Patients with recurrent sacrococcygeal pilonidal sinus disease.
Interventions	Intervention group: Off‐midline (Modified Karydakis flap) Comparator group: Off‐midline (Modified Limberg flap) Use of additional interventions in both groups: Drain: Not mentioned Methylene blue: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Not reported Secondary outcomes: Adverse effects (wound infection) Other outcomes reported by the study: Other adverse effects Follow‐up: Not mentioned
Notes	Publication status: Published

NCT01268969.

Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: Not reported Enrolment (Actual): 120 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Single (Participant) Other design features: Not reported
Participants	Location: Egypt Setting: Mansoura University Inclusion criteria for the study: People with pilonidal sinus Exclusion criteria: Pilonidal abscess Baseline characteristics: Age: "Child, adult, older adult" Gender: Not reported BMI: Not reported
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Healing Time points: Not reported Method of measurement: Not reported Secondary outcomes Early complication Time points: Not reported Method of measurement: Not reported Early return to work Time points: Not reported Method of measurement: Not reported Other outcomes reported by the study: Not reported Follow‐up: The follow‐up period ranged from 8 months to 2 years, with a mean follow‐up period of about 18 months.
Notes	Funding source: Mansoura University

NCT03765086.

Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: Not reported Actual sample size: 50 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Not blinded Other design features: Not reported
Participants	Location: Pakistan Setting: Services hospital Inclusion criteria for the study: Patients with pilonidal sinus of any sex and aged 20 years or above Non‐infected pilonidal sinus Exclusion criteria: Infected pilonidal sinus Uncontrolled diabetes Patient on immunosuppressive therapy Baseline characteristics: Age: 20 years and older Gender: All BMI: Not reported
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence Time points: 8 to 12 weeks Method of measurement: appearance of multiple discharging sinuses in operative site after surgical procedure Wound healing Time points: 4 to 6 weeks Method of measurement: duration of wound healing (complete epithelialisation of wound) Secondary outcomes: Not reported Other outcomes reported by the study: Percentage of patients developing postoperative pain Time points: 3 days Method of measurement: postoperative pain (measured by visual analogue score with 1 being minimum score and 10 being maximum score) after surgery
Notes	Funding source: Services Hospital, Lahore

NCT04017260.

Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: Not reported Actual sample size: 175 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Not blinded Other design features: Not reported
Participants	Location: Egypt Setting: Zagazig University Inclusion criteria for the study: PSD in intergluteal region Between 18 and 35 years old Fulfilling the diagnostic criteria of chronic discharging sinus/sinuses in natal cleft with or without surrounding tissue inflammation and with associated pain and bleeding on clinical evaluation Exclusion criteria: Recurrent pilonidal sinus Patients who were terminally ill, had uncontrolled diabetes, were immunocompromised or immunosuppressed Acute pilonidal abscess Baseline characteristics: Age: 18 to 40 years Gender: All BMI: Not reported
Interventions	Intervention group 1: Open and closed technique Intervention group 2: Limberg flap technique Intervention group 3: Karydakis flap technique Intervention group 4: Open technique Use of additional interventions in both groups: Antibiotic: All Drain: Not mentioned Methylene blue: All
Outcomes	Outcomes of interest in the review: Primary outcomes: Postoperative recurrence Time points: 1 year Method of measurement: Not reported Secondary outcomes: Operative time Time points: intraoperative Method of measurement: Not reported Infection postoperative Time points: 2 weeks Method of measurement: Not reported Other outcomes reported by the study: Not reported
Notes

NCT04572919.

Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: Not reported Actual sample size: 137 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Double (participant, care provider) Other design features: Not reported
Participants	Location: Egypt Setting: Zagazig University Hospitals Inclusion criteria for the study: Patients with non‐complicated pilonidal sinus 18 to 60 years old Exclusion criteria: Elderly patients with comorbid diseases Scars from a previous pilonidal surgery or recurrent cases Abscess presentations Diabetes mellitus, immunodeficiency, or other co‐morbidity American Society of Anesthesiologists physical status classification grade III‐IV Age < 18 years or > 60 years Baseline characteristics: Age: 18 to 60 Gender: All BMI: Not reported
Interventions	Intervention group: Lateral advancement flap Comparator group: Classic Limberg flap Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence: Time points: 12 to 36 months Method of measurement: Not reported Secondary outcomes: Not reported Other outcomes reported by the study: Not reported
Notes

Sabet 2016.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation method: Using convenient sampling method Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: Iran Setting: Hospital Sample size: 44 Number randomised: 24 to intervention (Limberg flap) 20 to comparator (Midline primary closure (tension‐free primary closure)) Number analysed: 24 to intervention (Limberg flap) 20 to comparator (Midline primary closure (tension‐free primary closure)) Baseline characteristics: Age (years, mean ± SD): Total: 15 ‐ 50 years Male (n, %): Not mentioned BMI: Not mentioned Type of presentation: Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: Not mentioned Control group: Not mentioned Inclusion criteria for the study: Patients with uncomplicated pilonidal sinus disease. Exclusion criteria for the study: Not mentioned
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Midline primary closure (tension‐free primary closure) Use of additional interventions in both groups: Drain: Not mentioned Methylene blue or Dye: Not mentioned Perioperative antibiotic: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Time to wound healing (days) Recurrence (2 years) Secondary outcomes: Length of hospital stay (days) Operative time (minutes) Other outcomes reported by the study: Disability score (1 month) Pain duration (days) Pain score (VAS, 1 month) Follow‐up: Patients were followed for two years.
Notes

Wright 2001.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: Not mentioned Blinding: Not mentioned CONSORT flow diagram: Not mentioned
Participants	Location: UK Setting: Hospital Sample size: Not mentioned Number randomised: Not mentioned Number analysed: Not mentioned Type of presentation: Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: Not mentioned Control group: Not mentioned Baseline characteristics: Age (years, mean ± SD): Not mentioned Male (n, %): Not mentioned BMI (mean ± SD): not mentioned Inclusion criteria for the study: Not mentioned Exclusion criteria for the study: Not mentioned
Interventions	Intervention group: Off‐midline (Bascom cleft lift) Comparator group: Midline primary closure Use of additional interventions in both groups: Not mentioned
Outcomes	Outcomes of interest in the review: Primary outcomes: Wound healing Secondary outcomes: Time to return to work (days) Other outcomes reported by the study: Not reported Follow‐up: The patients were then reviewed at 2‐weekly intervals until their wounds were healed. On each occasion, a further health assessment questionnaire was completed. Once healed, the patients were reviewed at 6‐monthly intervals for 2 years.
Notes	Poster. No data.

Youssef 2015.

Methods	Study design: Parallel‐group RCT Sample size calculation: Not mentioned Randomisation: A system of sequentially numbered sealed envelopes Blinding: Single‐blinded. No details CONSORT flow diagram: Not mentioned
Participants	Location: Egypt Setting: Hospital Sample size: 120 Number randomised: 60 to intervention (Modified Limberg flap) 60 to comparator (Tension‐free primary closure) Number analysed: 60 to intervention (Modified Limberg flap) 60 to comparator (Tension‐free primary closure) Baseline characteristics: Age (years, mean ± SD): Intervention group: 27.0 ± 6.8 Control group: 28.0 ± 7.6 Male (n, %): Total: 102, 85% Each group: Not mentioned BMI: Not mentioned Type of presentation Primary disease Intervention group: Not mentioned Control group: Not mentioned Recurrent disease Intervention group: Not mentioned Control group: Not mentioned Unspecified Intervention group: Not mentioned Control group: Not mentioned Inclusion criteria for the study: Patients with PSD were referred to Mansoura General and Colorectal Surgery Units. Exclusion criteria for the study: Participants are medically unfit or acute abscesses.
Interventions	Intervention group: Off‐midline (Modified Limberg flap) Comparator group: Midline primary closure (tension‐free primary closure) Use of additional interventions in both groups: Drain: Both yes Methylene blue: Not mentioned Perioperative antibiotic: Both yes. Perioperative antibiotics ‘ampicillin–sulbactam’ were used.
Outcomes	Outcomes of interest in the review: Primary outcomes: Recurrence rate Secondary outcomes: Adverse effects (wound infection, wound dehiscence) Length of hospital stay (days) Operative time (minutes) Postoperative or post‐procedural pain (VAS, 1st postoperative day) Quality of life (SF‐36, 3rd month) Other outcomes reported by the study: Seroma formation Operative blood loss (mL) Other adverse effects Patient satisfaction score Time to drain removal Time to sit on toilet and walk pain‐free Follow‐up: Every week for 1 month, monthly for the first year, then every 3 months Follow‐up period (mean ± SD): 43.5 ± 3.4, range 12 to 60 months
Notes

BMI: body mass index; PSD: pilonidal sinus disease; RCT: randomised controlled trial; SD: standard deviation; SF‐36: 36‐item short form quality of life questionnaire; SPD: sacrococcygeal pilonidal disease; VAS: visual analogue scale

Characteristics of ongoing studies [ordered by study ID]

IRCT201710109014N195.

Study name	Effect of simple closure versus Limberg flap on initial repair in patients with sacrococcygeal pilonidal sinus: a randomized clinical trial
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 76 Calculation method: Not reported Randomisation: Randomised Randomisation description: "The patients will be randomly assigned to intervention and control groups using block randomization." Blinding: Not blinded Other design features: Randomisation: "The patients will be randomly assigned to intervention and control groups using block randomization." Blinding: Not possible.
Participants	Location: Iran Setting: Besat Hospital Inclusion criteria for the study: Aged 18 to 65 years; sacrococcygeal pilonidal sinus Exclusion criteria: Infection of abscess in the area of pilonidal sinus; previous history of pilonidal sinus Baseline characteristics: Age: 18 to 65 years old Gender: Both BMI: Not reported
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Simple Midline closure Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Wound repair: Time points: 1 week and then 1, 3, and 23 months after surgery Method of measurement: through physical examination Secondary outcomes: Assessing pain Time points: 3 and 6 hours after surgery Method of measurement: VAS
Starting date	Registration date: 19 October 2017 Registration timing: Registered while recruiting Approval date: 17 June 2017 Recruitment status: Recruitment complete Expected recruitment start date: 22 June 2017 Expected recruitment end date: 20 March 2018 Actual recruitment start date: Not reported Actual recruitment end date: Not reported
Contact information	Person responsible for general inquiries Responsible person: Dr Abdollah Boostan Afrooz Street address: Besat Hospital, Shahed Square Phone: +98 81 3364 0030 Email: dr_boostan@yahoo.com Person responsible for scientific inquiries Responsible person: Dr Abdollah Boostan Afrooz Street address: Besat Hospital, Shahed Square Phone: +98 81 3364 0030 Email: dr_boostan@yahoo.com Person responsible for updating data Responsible person: Dr Jalal Poorolajal Street address: School of Public Health, Hamadan University of Medical Sciences, Shahid Fahmideh Ave Phone: +98 81 3838 0090 Email: poorolajal@umsha.ac.ir
Notes	Funding source Name: Vice‐chancellor for Research and Technology, Hamadan University of Medical Sciences. Type of organisation providing the funding: Academic Full name of funding responsible person: Dr Saeid Bashirian Grant name/code: Not reported

NCT03914729.

Study name	Multicenter randomized controlled trial of mobilized gluteus maximus muscle fascia flap versus primary closure in the treatment of primary and recurrent pilonidal sinus disease
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 84 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Double (participants, outcomes sssessor) Other design features: Not reported
Participants	Location: Russia Setting: Multicentre I.M. Sechenov First Moscow State Medical University Inclusion criteria for the study: Written informed consent Chronic primary or recurrent pilonidal sinus at the remission stage Presence or absence of secondary orifices Planned surgical treatment with excision of pilonidal sinus Location of secondary orifices less than 2 cm from the natal cleft The distance between bilateral symmetrical secondary orifices is less than 2 cm American Society of Anesthesiologists (ASA) score 1 to 3 Exclusion criteria: Acute pilonidal sinus abscess. The secondary opening (orifice) position is more than 2 cm from the midline ASA 4‐5 Predictable impossibility of following the protocol Pregnancy The patient is lost to further observation The patient refuses to continue to participate in the investigation Baseline characteristics: Age: 16 to 70 years old Gender: All BMI: Not reported
Interventions	Intervention group: Gluteus maximus fascia plasty flap Comparator group: Simple midline closure Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Wound repair: Time points: 6 months to 5 years Method of measurement: the rate of disease recurrence (clinical picture of pilonidal sinus and/or appearance of new openings in the intergluteal cleft and/or chronic unhealing wound and/or residual cavity in the wound area as confirmed by soft tissue ultrasound) Secondary outcomes: Operative time Time points: 1 day Method of measurement: the length of surgery in minutes Blood loss Time points: 1 day Method of measurement: the amount of blood lost during surgery Postoperative pain intensity ‐ early postoperative period Time points: on 1st, 3rd, 5th, and 7th postoperative day Pain intensity will be evaluated twice a day (in the morning and in the evening) with a patient‐reported VAS that ranges from 0 to 10 with 0 representing no pain and 10 representing intolerable pain. A total score will be recorded. Postoperative pain intensity ‐ late postoperative period Time points: on 10th, 14th, 21st, and 30th day after surgery Method of measurement: pain intensity will be evaluated once a day with a patient‐reported VAS that ranges from 0 to 10 with 0 representing no pain and 10 representing intolerable pain. A total score will be recorded. Surgical site infection rate Time points: 3 month after surgery Method of measurement: the rate of infectious inflammation of the wound as confirmed by the observing doctor In‐hospital stay Time points: 30 days Method of measurement: the duration of treatment after surgery until discharge from the hospital (in days) Wound haemorrhage rate Time points: within 30 days from surgery Method of measurement: the rate of haemorrhage from wound edges Wound seroma rate Time points: 90 days after surgery Method of measurement: the rate of seroma detection in the wound area as confirmed by soft tissue ultrasound Wound healing rate Time points: 6 months after surgery Method of measurement: the proportion of patients having their wound completely healed Wound healing speed Time points: 5 years after surgery Method of measurement: the time period between surgery and complete healing of the wound Secondary surgery rate Time points: 5 years after surgery Method of measurement: the rate of surgical procedures after initial surgery performed for recurrent disease and/or wound complications Patient satisfaction with cosmetic results Time points: 6 months, 1 year, 3 years, 5 years after surgery Method of measurement: patient‐reported with a scale from 0 to 10, where 0 corresponds to "completely unsatisfactory" and 10 corresponds to "completely satisfactory". A total score is registered. Overall quality of life Time points: 1 to 7 days before surgery, 1 month, 3 months, 1 year, 3 years, 5 years after surgery Method of measurement: assessed with patient‐reported questionnaire SF‐36. A total score in each of 8 sections will be calculated and transformed into a 0 to 100 scale with a score of 0 equivalent to maximum disability and a score of 100 equivalent to no disability.
Starting date	Registration date: 16 April 2019 Registration timing: Prospective Approval date: Not reported Recruitment status: Recruiting Expected recruitment start date: 30 December 2019 Expected recruitment end date: 30 December 2024 Actual recruitment start date: 20 April 2017 Actual recruitment end date: Not applicable Actual study completion date: Not applicable
Contact information	Contact: Darya Shlyk, MD + 7‐920‐520‐77‐06 shlyk@kkmx.ru Contact: Arcangelo Picciariello, MD +393492185104 picciariello@kkmx.ru
Notes	Funding source: Russian Society of Colorectal Surgeons

NCT04217824.

Study name	A prospective, randomized, double‐blind study. Comparison of Karydakis flap and Limberg flap treatment of pilonidal sinus in adolescent [sic]
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 90 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Double (participants, outcomes assessor) Other design features: Not reported
Participants	Location: Turkey Setting: Kahramanmaras Sutcu Imam University Inclusion criteria for the study: Pilonidal sinus disease Exclusion criteria: Disease other than pilonidal sinus disease Baseline characteristics: Age: 10 to 18 years Gender: All BMI: Not reported
Interventions	Intervention group: Off‐midline (Karydakis flap) Comparator group: Off‐midline (Limberg flap) Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Postoperative recurrence Time points: 1 month Method of measurement: pain, swelling, and redness upon examination Secondary outcomes: Postoperative infection Time points: 1 week Method of measurement: purulent drainage Postoperative wound dehiscence Time points: 1 week Method of measurement: wound separation
Starting date	Registration date: 3 January 2020 Registration timing: Registered while recruiting Approval date: Not reported Recruitment status: Recruiting Expected recruitment start date: 22 May 2019 Expected recruitment end date: 22 May 2021 Actual recruitment start date: 22 May 2019 Actual recruitment end date: Not reported Actual study completion date: Not reported
Contact information	Contact: Ahmet Gokhan Guler +905057702633 drgokhanguler@gmail.com Contact: Ali Erdal Karakaya +905326313017 dralierdal@yahoo.com
Notes

NCT05289297.

Study name	Karydakis flap procedure versus Burow's triangle advancement flap procedure in the surgical treatment of sacrococcygeal pilonidal sinus disease
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 80 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Single (participant) Other design features: Not reported
Participants	Location: Turkey Setting: Istanbul Sultanbeyli State Hospital Inclusion criteria for the study: Patients between the ages of 18 and 65 Primary pilonidal sinus disease No abscess and absence of active infection during the operation Exclusion criteria: Patients under the age of 18 and over the age of 65 Recurrent pilonidal sinus cases Patients with chronic comorbidities such as immunosuppression, collagen tissue disease, insulin‐dependent diabetes mellitus, and neurological disease Patients with an ASA score of 3‐4 Patients with a body mass index greater than 35 kg/m2 Patients with drug and alcohol addiction Baseline characteristics limitation: Age: 18 to 65 years Gender: All BMI: Not reported
Interventions	Intervention group: Karydakis flap procedure Comparator group: Burow's triangle advancement flap procedure Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Pain score according to the visual analogue scale Time points: 1 to 3 days Method of measurement: measured by the visual analogue score with 1 being minimum score and 10 being maximum score Postoperative wound healing time Time points: 4 to 6 weeks Method of measurement: number of weeks to complete epithelialisation of the wound Procedure‐related complications Time points: 1 to 6 weeks Method of measurement: number of complications related to the surgery; type (wound dehiscence, seroma, haematoma, surgical site infection) Time to return to normal activity Time points: 4 to 6 weeks Method of measurement: the number of weeks to patient returning to normal activity Secondary outcomes: Recurrence of pilonidal sinus Time points: 6 to12 months Method of measurement: not reported
Starting date	Registration date: 21 March 2022 Registration timing: Prospectively Approval date: Not reported Recruitment status: Recruiting Expected recruitment start date: Not reported Expected recruitment end date: 1 June 2023 Actual recruitment start date: 12 June 2022 Actual recruitment end date: Not applicable Actual study completion date: Not applicable
Contact information	Contact: Muhammer Ergenç, MD 902165642400 ext 1321 muhammerergenc@gmail.com
Notes

NL4763.

Study name	Which technique for closing the wound after excision of a pilonidal sinus leads most quickly to complete wound healing? Primary closure in the midline and Limberg plastic are examined
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 74 Calculation method: Not reported Randomisation: Randomised Randomisation description: Not reported Blinding: Single‐blinded (masking used) Other design features: Not reported
Participants	Location: Netherlands Setting: The Medisch Spectrum Twente Inclusion criteria for the study: Primary or recurrent pilonidal sinus Width of fistula complex between 1 cm and 3 cm Signed informed consent Age: >/= 18 years Mentally competent Exclusion criteria: Patients with an actual abscess Not able to follow up Patients in which wound healing problems are to be expected (diabetes type II, immunosuppression, chemotherapy, HIV/AIDS) Not mentally competent Dementia Baseline characteristics: Age: >/= 18 years Gender: Not mentioned BMI: Not mentioned
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Simple midline closure Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Wound healing time Time points: Not reported Method of measurement: Not reported Secondary outcomes: Wound infections Time points: Not reported Method of measurement: Not reported Recurrence Time points: 1 year Method of measurement: Not reported Follow‐up: Wound check after 7 to 10 days; wound photo; complete form 3/VAS score Wound check after 14 to 21 days (in principle remove stitches); wound photo; complete form 3/VAS score Control after 6 weeks; wound photo; complete form 3/EQ5‐D (EuroQoL quality of life instrument) Control after 3 months; wound photo; fill out form 3/EQ5‐D (Telephone) check after 6 months; fill out form 3 (Telephone check after 12 months; complete form 3 Any additional outpatient checks
Starting date	Registration date: 04 November 2014 Registration timing: Prospectively Approval date: Not reported Recruitment status: Completed Expected recruitment start date: 01 January 2015 Expected recruitment end date: Not reported Actual recruitment start date: Not reported Actual recruitment end date: Not reported Actual study completion date: 01 January 2016
Contact information	J.M. Klaase +31 (0)53 48720 00
Notes

NL4792.

Study name	Limberg Trial
Methods	Study design: Parallel‐group RCT Sample size calculation: Target sample size: 74 Calculation method: Not reported Randomisation: Randomized Randomization description: Not reported Blinding: Single‐blinded (masking used) Other design features: Not reported
Participants	Location: Netherlands Setting: Medisch Spectrum Twente Inclusion criteria for the study: Primary or recurrent pilonidal sinus Width of fistula complex between 1 cm and 3 cm Signed informed consent Age: >/= 18 years Mentally competent Exclusion criteria: Patients with an actual abscess Not able to follow up Patients in which wound healing problems are to be expected (diabetes type II, immunosuppression, chemotherapy, HIV/AIDS) Not mentally competent Dementia Baseline characteristics: Age: >/= 18 years Gender: Not mentioned BMI: Not reported
Interventions	Intervention group: Off‐midline (Limberg flap) Comparator group: Simple Midline closure Use of additional interventions in both groups: Not mentioned
Outcomes	Primary outcomes: Time to complete wound healing Time points: Not reported Method of measurement: Not reported Secondary outcomes: Wound infections Time points: Not reported Method of measurement: Not reported Recurrence Time points: 1 year Method of measurement: Not reported
Starting date	Registration date: 18 November 2014 Registration timing: Prospectively Approval date: Not reported Recruitment status: Completed Expected recruitment start date: 01 January 2015 Expected recruitment end date: Not reported Actual recruitment start date: Not applicable Actual recruitment end date: 01 January 2016
Contact information	Medisch Spectrum Twente Postbus 50000 A. Stam Enschede 7500 KA The Netherlands researchbureau.heelkunde@ziekenhuis‐mst.nl
Notes

ASA: American Society of Anesthesiologists; BMI: body mass index; CI: confidence interval; RCT: randomised controlled trial; VAS: visual analogue scale

Differences between protocol and review

Methods: types of participants

• "Studies that included participants presenting with an abscess were excluded, except when data could be extracted and analysed separately for individuals with and without pilonidal abscesses." An incorrect description in the protocol was revised.

Methods: types of interventions

• In the protocol, we initially considered "all types of midline closure (e.g. tension‐free midline closure)". However, during the review process, we encountered numerous comparisons and clinical heterogeneity amongst the studies. Due to the limited length for this review and clinical heterogeneity amongst studies, we revised this section as "midline closure (including conventional procedures and tension‐free procedures without flap techniques)"

Methods: types of outcome measures

• We revised the outcome "The proportion of people whose wounds have completely healed at a given time point, assessed clinically by researchers or the clinical team" to "Proportion of wounds healed (number of wounds healed/not healed) during short‐ (30‐day), medium‐ (6‐month) and long‐term (1‐year) follow‐up, assessed clinically by researchers or a clinical team".

• In the protocol, we initially included "all reported adverse events" as an outcome of interest. However, during the review process, we decided to exclude this outcome due to the inconsistency in the definition of total adverse events across the studies. To ensure the validity and reliability of the results, we only considered adverse events (wound infection and wound dehiscence) that were clearly defined and consistently reported across all included studies.

• We revised the outcome "Time to return (in days) to normal daily activities" as "Time to return to work (days)" to align with another Cochrane Review (Al‐Khamis 2010). Due to the limited number of studies reporting "Time to return (in days) to normal daily activities", we found it necessary to adapt our approach accordingly.

• We observed variation in the time points for the outcome "Postoperative or post‐procedural pain" amongst the included studies. Consequently, we restricted our analysis to postoperative or post‐procedural pain reported on postoperative day 1.

Methods: dealing with missing data

• For dichotomous outcomes (e.g. recurrence), our initial plan was to calculate the effect estimate using the number of participants randomised in the group as the denominator. However, we revised this approach and instead analysed only the available data (i.e. disregarding the missing data) and assessed the risk of bias due to missing outcome data in randomised trials.

• We replaced "If SDs could not be calculated, we planned to impute them; for example, from other studies in the meta‐analysis" with "In cases where sample means and SDs could not be calculated, we estimated these figures using the five-number summary (Luo 2018, Wan 2014). In the absence of available mean values, we utilised median values as a substitute."

Methods: assessment of reporting biases

• "We used Egger's test to determine the statistical significance of reporting bias (Egger 1997), and considered P < 0.05 to be a statistically significant reporting bias" (previously not planned).

Methods: subgroup analysis and investigation of heterogeneity

• "We planned to carry out the following subgroup analyses for the outcomes included in the summary of findings tables" (previously "primary outcomes").

• "At least three studies are available" (previously "at least 10 studies").

Methods: sensitivity analysis

• "If we were able to include at least three studies for the outcomes presented in the summary of findings tables" (previously "primary outcomes" and "at least 10 studies").

• "We repeated the analysis excluding studies that did not provide clear tension‐free method descriptions for the comparison of off‐midline closure versus tension‐free midline closure" (added post‐protocol).

Methods: summary of findings and assessment of the certainty of the evidence

• In our initial protocol, we intended to prepare a summary of findings table for each comparison. However, in the review, we focused on the following comparisons:

  • Off‐midline closure versus conventional midline closure

  • Off‐midline closure versus tension‐free midline closure

  • Off‐midline (Karydakis flap) versus off‐midline (Limberg flap)

Contributions of authors

Zhaolun Cai: conceived the review; designed the review; coordinated the review; extracted data; checked quality of data extraction; analysed or interpreted data; undertook quality assessment; checked quality assessment; performed statistical analysis; checked quality of statistical analysis; produced the first draft of the review; contributed to writing or editing the review; advised on the review; wrote to study authors; approved final review prior to submission; is guarantor of the review.

Zhou Zhao: conceived the review; designed the review; checked quality of data extraction; analysed or interpreted data; checked quality assessment; checked quality of statistical analysis; produced the first draft of the review; contributed to writing or editing the review; advised on the review; performed translations; wrote to study authors; approved final review prior to submission.

Qin Ma: conceived the review; designed the review; produced the first draft of the review; contributed to writing or editing the review; advised on the review; approved final review prior to submission.

Chaoyong Shen: produced the first draft of the review; contributed to writing or editing the review; approved final review prior to submission.

Zhiyuan Jiang: produced the first draft of the review; contributed to writing or editing the review; advised on the review; approved final review prior to submission.

Chunyu Liu: produced the first draft of the review; contributed to writing or editing the review; approved final review prior to submission.

Chunjuan Liu: produced the first draft of the review; contributed to writing or editing the review; approved final review prior to submission.

Bo Zhang: conceived the review; designed the review; coordinated the review; extracted data; checked quality of data extraction; analysed or interpreted data; undertook quality assessment; checked quality assessment; performed statistical analysis; checked quality of statistical analysis; produced the first draft of the review; contributed to writing or editing the review; advised on the review; wrote to study authors; performed translations; approved final review prior to submission; is guarantor of the review.

Sources of support

Internal sources

• No sources of support provided

External sources

• The National Institute for Health Research (NIHR), UK

  This project was supported by the National Institute for Health Research, via Cochrane Infrastructure funding to Cochrane Wounds. The views and opinions expressed are those of the authors and do not necessarily reflect those of the NIHR, NHS or the Department of Health.

Declarations of interest

Zhaolun Cai: works as a healthcare professional.

Zhou Zhao: works as a healthcare professional.

Qin Ma: works as a healthcare professional.

Chaoyong Shen: works as a healthcare professional.

Zhiyuan Jiang: none known.

Chunyu Liu: works as a healthcare professional.

Chunjuan Liu: none known.

Bo Zhang: works as a healthcare professional.

These authors contributed equally to this work.

New